Some embodiments of the present disclosure provide an image fusion method, an apparatus, an electronic device, and a readable storage medium. The method includes: acquiring at least two RAW images to be processed in a same scenario; regarding one of the at least two RAW images to be processed as a reference frame, the others as supplementary frames, and determining a luminance relationship between each of the supplementary frames and the reference frame; for each of the supplementary frames, linearly adjusting the luminance of pixels in the supplementary frame based on the luminance relationship to obtain an adjusted supplementary frame; and fusing each of the adjusted supplementary frames and the reference frame to obtain a fused image.
The present disclosure provides a temperature correcting method, apparatus and system, which relates to the technical field of computers. The method includes: acquiring a target reference image-frame pair containing a person, wherein the target reference image-frame pair includes a target visible-light image frame and a target thermal-imaging image frame; converting the target visible-light image frame into a depth image frame; inputting the depth image frame and the target thermal-imaging image frame into an atmosphere-inverse-scattering model, wherein the atmosphere-inverse-scattering model refers to a neural network model obtained by fitting a relation between depths and temperature correcting values in advance; and by using the atmosphere-inverse-scattering model, based on the depth image frame, performing temperature correction to the target thermal-imaging image frame, to obtain a target temperature. The present disclosure may effectively increase the accuracy of person temperature measurement.
A warehousing system, a shuttle vehicle for the warehousing system, and a navigation method therefor. The shuttle vehicle is provided with a memory (110), a communication apparatus (120), a positioning sensor (130), a camera apparatus (140), and a controller (150). The memory (110) is used to store a topological map of the warehousing system, each node of the topological map comprising location identification information of a corresponding parking space in the warehousing system. The communication apparatus (120) is used to receive a traveling route instruction, the traveling route instruction comprising information of a traveling route from a starting point to a destination. The positioning sensor (130) is used to acquire current location information of the shuttle vehicle during traveling of the shuttle vehicle. The camera apparatus (140) is used to photograph an image of the warehousing system during traveling of the shuttle vehicle. The controller (150) is used to control the shuttle vehicle to travel from the starting point to the destination on the basis of the topological map and the information of the traveling route.
The present application relates to a parallax image processing method, an apparatus, a computer device and a storage medium. The method includes: obtaining at least one mask image by performing down-sampling on an initial parallax image, and then performing patch processing and pixel replacement processing on the mask image to obtain a processed mask image, and determining an abnormal parallax value on the initial parallax image according to the processed mask image, and finally performing interpolation processing and filtering processing on the abnormal parallax value on the initial parallax image to obtain a target parallax image.
The present application belongs to the field of logistics management, and relates to a warehousing system, a goods shelf management method, apparatus and system, a control terminal, and a readable storage medium. The warehousing system comprises a plurality of goods shelf groups, a control terminal, and a handling robot. Each goods shelf group comprises a plurality of goods shelves which are arranged in a horizontal direction, and a mobile apparatus is mounted at the bottom of each goods shelf. The control terminal is configured to control the goods shelves to move, so as to open a channel at a target goods shelf. The handling robot is configured to move to the channel at the target goods shelf, so as to execute an outbound task or an inbound task. The goods shelf management method comprises: determining, according to the received current task, a target goods shelf corresponding to the current task; determining whether a channel at the target goods shelf is in an opened state; and when the channel at the target goods shelf is in an unopened state, controlling at least one goods shelf in a goods shelf group where the target goods shelf is located to move, so as to open the channel at the target goods shelf. By means of the warehousing system and the goods shelf management method, the utilization rate of a horizontal space of a warehouse can be improved, which is conducive to reducing the cost of unit storage of the warehouse.
G06Q 10/04 - Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
The present disclosure provides a temperature measurement method, a temperature measurement device, an electronic apparatus and a computer-readable storage medium. The method includes obtaining an image frame pair including a target object by a visible light camera and a thermal imaging camera, and a blackbody being also set in an image acquisition region of the thermal imaging camera; determining a measured temperature of the target object based on the image frame pair; performing a blackbody detection on the infrared image to obtain a detection result of the blackbody; determining a measured temperature of the blackbody based on the detection result of the blackbody and the infrared image; and correcting the measured temperature of the target object according to the measured temperature of the blackbody and a preset temperature of the blackbody, a corrected temperature being used as a temperature measurement result of the target object.
Embodiments of the present disclosure provide an image stitching method and an electronic device, the method comprising: acquiring a first image and a second image to be stitched; determining an initial stitched image of the first image and the second image; by using a first neural network model, fusing a target overlapping area in the initial stitched image to obtain a fused overlapping area corresponding to the target overlapping area; and, on the basis of the fused overlapping area and the initial stitched image, determining a target stitched image corresponding to the first image and the second image. In said method, after determining the initial stitched image of the first image and the second image, the first neural network model is used to fuse the target overlapping area in the initial stitched image to obtain a corresponding fused overlapping area, and it is not necessary to perform fusion-related computation on each pixel in the initial stitched image on the basis of a CPU, reducing the time for fusion and computation on all pixels in the initial stitched image, improving fusion efficiency, and thus improving the processing efficiency of image stitching.
A multi-camera zoom control method and apparatus, and an electronic system and a storage medium. The method includes: in the process of a first camera collecting an image, if the current set magnification input by a user is in a magnification transition zone, starting a second camera; acquiring a corresponding stereo correction matrix on the basis of calibration parameters of the first camera and the second camera; calculating a translation matrix on the basis of an acquired pixel position corresponding relationship between the same content regions of interest that correspond to a first zoomed image and a second zoomed image and in combination with the current set magnification, and then calculating a smooth transition transformation matrix in combination with the stereo correction matrix; and performing, by applying the smooth transition transformation matrix, affine transformation on an image output by the first camera, to obtain a display image of a device.
H04N 5/262 - Studio circuits, e.g. for mixing, switching-over, change of character of image, other special effects
H04N 13/239 - Image signal generators using stereoscopic image cameras using two 2D image sensors having a relative position equal to or related to the interocular distance
An image enhancement method and an electronic device. The method comprises: obtaining at least two images in a target scene to be processed, and processing the brightness of a first image to be processed, to obtain a second image to be processed; performing first fusion processing on said first image and said second image to obtain a fused image to be processed; and performing second fusion processing on said fused image and the images in said at least two images other than said first image to obtain an enhanced fused image. The first fusion processing is performed on said first image and said second image, such that a good image effect of said obtained fused image is achieved, thereby achieving a good image effect of the enhanced fused image obtained on the basis of said fused image having a good image effect.
Provided in the present disclosure are a three-dimensional reconstruction method and apparatus, and an electronic device and a storage medium. The method comprises: acquiring a plurality of pieces of monitoring image data, wherein the plurality of pieces of monitoring image data each comprise image data of a vehicle; determining a target image group of a vehicle of a target model according to the plurality of pieces of monitoring image data, wherein the target image group comprises images of the vehicle of the target model at different viewing angles; acquiring a calibration result of a camera corresponding to each image in the target image group; and obtaining a three-dimensional model of the vehicle of the target model according to the target image group and the calibration result of the camera corresponding to each image in the target image group. By means of the method, there is no need to manually hold a three-dimensional scanning device to scan a vehicle, and a three-dimensional vehicle model can be constructed on the basis of monitoring image data, such that the efficiency is higher; moreover, there is no need to use the three-dimensional scanning device and a physical vehicle, such that the cost is lower, and the implementation is easier.
Provided by the present application are a pose determination method and apparatus, an electronic device, and a readable storage medium. A specific embodiment of the method comprises: obtaining an image set to be processed, the image set to be processed comprising a plurality of target images; for each of the target images, on the basis of target key point information of the target image, determining initial pose information of a target object; and on the basis of an initial model and the initial pose information corresponding to each of the plurality of target images, determining a target model and a target pose. In the described method, a target model and target pose can be determined at the same time, so that there is no need to generate an object model in advance and no need to design a matching algorithm between the target object and the object model, optimizing the processing steps of a pose estimation process.
The present disclosure provides an attitude determining method, an electronic device, and a readable storage medium. The attitude determining method comprises: acquiring an image to be processed, the image to be processed comprising an image of a target vehicle; determining, on the basis of historical yaw angle information of the target vehicle, an estimated attitude of the target vehicle at the time of capturing the image to be processed, the historical yaw angle information being yaw angle information of the target vehicle at a time before the time of capturing the image to be processed; verifying whether the estimated attitude satisfies a verification condition; and if yes, determining the estimated attitude as an actual attitude of the target vehicle at the time of capturing the image to be processed. According to the attitude determining method of the present disclosure, the actual attitude currently corresponding to the target vehicle can be quickly determined, and the number of paths of system multiplexing is increased.
Provided in the embodiments of the present application are a de-pixelating method, an electronic device and a storage medium. The method comprises: acquiring a target RGBIR image; and performing first processing on the target RGBIR image by using a first neural network, so as to obtain a target IR image, and/or, performing second processing on the target RGBIR image by using a second neural network, so as to obtain a target RGB image, wherein the first processing comprises: eliminating a visible light component from a pixel value that corresponds to an IR-channel pixel point in the target RGBIR image, and predicting a pixel value of a color-channel pixel point in an IR channel; and the second processing comprises: for each color channel, eliminating an infrared radiation component from a pixel value corresponding to the color-channel pixel point in the target RGBIR image, and predicting a pixel value of a pixel point in the color channel, other than the color-channel pixel point.
The present application provides an action recognition method and apparatus and an electronic device. The method includes: if a target object is detected from a video frame, acquiring a plurality of images containing the target object, and optical-flow images of the plurality of images; extracting an object trajectory feature of the target object from the plurality of images, and extracting an optical-flow trajectory feature of the target object from the optical-flow images of the plurality of images; and according to the object trajectory feature and the optical-flow trajectory feature, recognizing a type of an action of the target object. Because it combines the time-feature information and the spatial-feature information of the target object, effectively increases the accuracy of the detection and recognition on the action type, and may take into consideration the detection efficiency at the same time, thereby improving the overall detection performance.
Provided in the present disclosure are a splicing seam search method and apparatus for a video image, and a video image splicing method and apparatus. The splicing seam search method comprises: acquiring an energy map of each frame of video image in a first video; for a first frame of video image, determining a splicing seam search result thereof on the basis of the energy map thereof; for each of the remaining frames of video images, determining the range of a splicing seam search area on the basis of a splicing seam search result of the previous frame of video image; and on the basis of an energy map of the current video image, determining a splicing seam search result thereof within the range. In this way, a splicing seam search result is determined on the basis of an energy map of a video image; moreover, for a video image other than a first frame, the range of a splicing seam search area is first determined on the basis of a splicing seam search result of the previous frame of video image, and a splicing seam search result is then determined within the range of the splicing seam search area. By means of constraining the range of the splicing seam search area, the difference between splicing seam areas of two consecutive frames of video images can be reduced, such that the problem of jittering of a spliced video during a playing process is alleviated, thereby improving the splicing effect of a panoramic video.
Provided are an image rectification method and apparatus, and an electronic system. The image rectification method includes: acquiring a first image and a second image of the same shooting object by means of a first shooting apparatus and a second shooting apparatus which are coaxially disposed; and correcting the second image according to shooting parameters of the first shooting apparatus and the second shooting apparatus to obtain a second rectified image, such that the parallax between the second rectified image and the first image in a vertical direction or a horizontal direction is zero. In the method, by taking a first image as a reference, and by means of adjusting the shooting parameters of the first shooting apparatus and a second shooting apparatus, only the second image is rectified, thereby improving the operation efficiency of image rectification, and improving the accuracy and stability of an image rectification result.
G06V 10/44 - Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersectionsConnectivity analysis, e.g. of connected components
G06V 10/74 - Image or video pattern matchingProximity measures in feature spaces
G06V 10/98 - Detection or correction of errors, e.g. by rescanning the pattern or by human interventionEvaluation of the quality of the acquired patterns
17.
TARGET RETRIEVAL METHOD AND APPARATUS, DEVICE AND STORAGE MEDIUM
The present application provides a target retrieval method and apparatus, an electronic device, and a storage medium. The method comprises: performing target detection on an image under retrieval by means of a detection model corresponding to a target to be retrieved to obtain at least one detection frame having a confidence greater than or equal to a confidence threshold, the confidence threshold being lower than a threshold for a conventional detection network, and the detection model corresponding to the target being a model obtained by performing parameter updating on the basis of a template image of the target; and filtering the at least one detection frame by means of a post-processing model to obtain a detection frame corresponding to the target. In the present application, because the detection model is a model obtained by performing parameter updating on the basis of the template image of the target, a pre-trained detection model is updated by using a small quantity of template images, and the confidence threshold is low, it can be ensured that sufficient detection frames are recalled, and the detection frame corresponding to the target to be retrieved is obtained by filtering by means of the post-processing model, thereby achieving the retrieval of a small sample target.
The present disclosure provides a distance determination method, apparatus and system, relating to the technical field of image processing. The method includes the following steps: acquiring a master visual image photographed by a master camera and an original auxiliary visual image photographed by an auxiliary camera; acquiring an initial matching point pair between the master visual image and the original auxiliary visual image through feature extraction and feature matching; correcting the original auxiliary visual image sequentially, based on the initial matching point pair and different constraints, so as to obtain a target auxiliary visual image, wherein the different constraints includes: a constraint of a minimum rotation angle and a constraint of a minimum parallax; and determining a focusing distance according to the master visual image and the target auxiliary visual image. The focusing distance can be determined more accurately.
G06V 10/74 - Image or video pattern matchingProximity measures in feature spaces
G06V 10/24 - Aligning, centring, orientation detection or correction of the image
G06V 10/44 - Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersectionsConnectivity analysis, e.g. of connected components
G06V 10/98 - Detection or correction of errors, e.g. by rescanning the pattern or by human interventionEvaluation of the quality of the acquired patterns
G06T 7/593 - Depth or shape recovery from multiple images from stereo images
The present application provides a camera calibration method and apparatus, an electronic device, and a storage medium. The method comprises: obtaining information of bounding boxes of areas where at least two targets in a first image are located, wherein the first image is an image captured by a camera at a first moment, and the information of each bounding box comprises the position and the size of the corresponding bounding box in the first image; determining a first linear parameter group according to the information of the bounding boxes of the at least two targets, the first linear parameter group being used for representing a linear relationship between the position and the size of the bounding box in the first image; when the first linear parameter group is present in a predetermined correspondence between a linear parameter group and a camera parameter group, finding a first camera parameter group corresponding to the first linear parameter group from the correspondence, the first camera parameter group comprising the mounting height of the camera, the pitch angle of the camera, and the focal length of the camera; and determining a target camera parameter group according to the first camera parameter group. Camera calibration can be completed without placing a calibration object or controlling the camera to move.
Provided in the present disclosure are an object identification method and apparatus, and an electronic system. The method comprises: acquiring a first image, which includes a target object; if visible parts of the target object in the first image do not include all parts of the target object, performing deformation and filling processing on the first image, such that relative positions, in the first image, of the visible parts of the target object included in the first image meet a specified standard; and then, extracting an object feature of the target object from the processed first image, so as to identify the target object. In the method, when the parts of a target object in an image are incomplete, deformation and filling processing is performed on the image, such that relative positions, in the image, of the parts in the target object match the relative positions thereof when the image includes all the parts of the target object; and the target object can be identified by means of directly extracting an object feature of the target object from the processed image, without the need to perform local segmentation and identification on each part, thereby reducing the calculation complexity of object identification, and facilitating large-scale deployment.
The present disclosure relates to the technical field of image processing, and provided in embodiments thereof are a target detection method, an apparatus, an electronic device, and a storage medium, which can reduce the size of an image without needing to use a means of coding and decoding and without affecting the detection performance on a target object in a size-reduced image. The target detection method may comprise: performing color coding on an input original color image, and obtaining a plurality of YUV color space images; performing pixel region division on a target image among the plurality of images, and obtaining a plurality of pixel regions corresponding to the target image; performing discrete cosine transform on each pixel region, and obtaining transform features of the target image; selecting a target channel of a low frequency region among the transform features; and performing object detection according to frequency domain feature information of the target channel in the target image.
The present application provides an image recognition method and apparatus, an electronic device, and a storage medium. The method comprises: obtaining an image to be recognized of an object to be recognized; obtaining recognition information according to said image and a template image, the recognition information comprising a similarity score between the image to be recognized and the template image, and the number of matching points and/or an overlap area between the image to be recognized and the template image; and obtaining a recognition result according to the recognition information. The solution provided by the present application improves the recognition accuracy of an image to be recognized such as a fingerprint or palmprint image.
The present application provides an obstacle avoidance method and apparatus and an automatic navigation apparatus, which are applied in the field of robots. The obstacle avoidance method comprises: acquiring the motion state of an automatic navigation apparatus and obstacle information on a driving route of the automatic navigation apparatus, wherein the obstacle information comprises an obstacle motion state and obstacle attributes; and according to the motion state of the automatic navigation apparatus and the obstacle information, determining an obstacle avoidance strategy corresponding to the automatic navigation apparatus. In the solution above, during the obstacle avoidance process, the automatic navigation apparatus will comprehensively take into consideration the motion state thereof, as well as the obstacle motion state and the obstacle attributes, so that obstacle avoidance is more accurately performed to avoid hazards.
An image processing method, an electronic device, and a computer readable storage medium, relating to the field of image processing. The method comprises: obtaining face key point information of an image to be processed (S110); according to the face key point information, determining a movement reference point when moving pixels in a face area (S120), and dividing the face area into different local areas (S130); and during face thinning, moving pixels included in the local areas toward the movement reference point according to pixel moving policy parameters corresponding to the local areas to which the pixels belong (S140). Therefore, the final face thinning effect can be natural as much as possible, thereby improving the visual effect after face thinning.
The present disclosure provides a zoom control method and apparatus, an electronic device, and a computer-readable storage medium. The method comprises: when a magnification inputted by a user is in a switching magnification interval, by means of a matching relationship between a first image and a second image, determining parameters such as a translation amount between the images and a relative positional relationship between cameras, and processing the first image on the basis of said parameters; and at the same time, determining a switching magnification by means of the focal length of a camera and the scale of the images, so that when the magnification inputted by the user reaches the switching magnification, the first image and second image after processing have the same scale and field of view, thus achieving the smooth switching between displayed images and improving the zoom effect. In addition, since the method does not require the calibration of parameters, a smooth and stable zoom effect may still be ensured when modules age or collide.
An obstacle bypassing method for a mobile device, and a mobile device and a storage medium. The method comprises: on the basis of a path planning instruction, determining an initial path from the current position to an obstacle avoidance position, wherein the obstacle avoidance position is located on a global path (S101); sampling the initial path, and taking a path sampling point as an initial position of a control point of a curve to be fitted (S102); optimizing the position of the control point on the basis of the initial position, the distance between the control point and an obstacle, and a speed parameter of a mobile device at the control point (S103); and according to the optimized position of the control point, determining a local target path from the current position to the obstacle avoidance position (S104). The position of a control point is taken as an optimization variable, and compared with an optimization-based method, optimization parameters are reduced, the time taken for finding a solution is shortened, and a real-time requirement is met on a low-cost calculation unit.
The present application relates to the technical field of image processing, and provides an image splicing method and apparatus, and a storage medium and an electronic device. The method comprises: acquiring a first image and a second image; calculating a first intermediate optical flow and a second intermediate optical flow according to the first image and the second image, wherein the first intermediate optical flow is an optical flow between an intermediate image and the first image, the second intermediate optical flow is an optical flow between the intermediate image and the second image, the intermediate image is an image at a viewing angle between those of the first image and the second image, the size of the first intermediate optical flow is the same as that of the first image, and the size of the second intermediate optical flow is the same as that of the second image; and calculating a spliced image of the first image and the second image according to the first intermediate optical flow, the second intermediate optical flow, the first image and the second image. By means of the image splicing method, image splicing can be completed without the need of calculating a plurality of homography matrices by means of complex iterations, and therefore the efficiency of image splicing can be improved.
The embodiments of the present application provide a keypoint detection method, an electronic device, a program, and a storage medium. The method comprises: by means of a backbone network, extracting image features of an image to be detected, and obtaining a feature map, the image to be detected comprising a target object; performing keypoint detection on the feature map according to multiple pose keypoint templates, and obtaining at least one set of keypoints corresponding to the target object, the pose keypoint templates representing relative positional relationships between multiple keypoints corresponding to a pose; screening the at least one set of keypoints, and obtaining a keypoint detection result of the target object. In the present application, keypoint detection results of all target objects in an image to be detected can be directly determined on the basis of multiple pose keypoint templates, thus increasing efficiency in keypoint detection.
Provided in the present application are a traveling route planning method and apparatus, a storage medium, and a mobile apparatus. The method comprises: acquiring a traveling route and a configuration file of an automated navigation mobile apparatus within a calibration area, wherein the configuration file comprises a motion restriction condition for the automated navigation mobile apparatus at each position in the calibration area; and verifying, according to the configuration file, whether the motion restriction condition for a corresponding position is met when the automated navigation mobile apparatus travels according to the traveling route, and determining a final traveling route according to a verification result. By means of the technical solution of the present application, the safety of the generated final traveling route can be improved.
G06K 17/00 - Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups , e.g. automatic card files incorporating conveying and reading operations
30.
METHOD AND DEVICE FOR IMAGE FUSION, COMPUTING PROCESSING DEVICE, AND STORAGE MEDIUM
The present application relates to an image-fusion method and apparatus, a computing and processing device and a storage medium. The method includes: based on a same one target scene, acquiring a plurality of exposed images of different exposure degrees; acquiring a first exposed-image fusion-weight diagram corresponding to each of the exposed images, wherein the first exposed-image fusion-weight diagram contains fusion weights corresponding to pixel points of the exposed image; acquiring a region area of each of overexposed regions in each of the exposed images; for each of the exposed images, by using the region area of each of the overexposed regions in the exposed image, performing smoothing filtering to the first exposed-image fusion-weight diagram corresponding to the exposed image, to obtain a second exposed-image fusion-weight diagram corresponding to the exposed image; and according to each of the second exposed-image fusion-weight diagrams, performing image-fusion processing to the plurality of exposed images, to obtain a fused image. Accordingly, the present application can balance the characteristics of the different overexposed regions, and prevent the losing of the details of the small overexposed regions, to enable the obtained fused image to be more realistic.
G06V 10/75 - Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video featuresCoarse-fine approaches, e.g. multi-scale approachesImage or video pattern matchingProximity measures in feature spaces using context analysisSelection of dictionaries
G06V 10/22 - Image preprocessing by selection of a specific region containing or referencing a patternLocating or processing of specific regions to guide the detection or recognition
31.
DISPARITY MAP OPTIMIZATION METHOD AND APPARATUS, AND ELECTRONIC DEVICE AND COMPUTER-READABLE STORAGE MEDIUM
The present application relates to the field of image processing and relates to a disparity map optimization method and apparatus, and an electronic device and a computer-readable storage medium. The method comprises: obtaining a first image, a second image, and an initial disparity map corresponding to the first image and the second image; according to the initial disparity map, the first image, and the second image, determining an energy equation used for guiding filtering, the energy equation comprising a data item and a smooth item; and performing guiding filtering on the initial disparity map according to the energy equation, and obtaining an optimized disparity map. By means of the method, errors can be prevented from diffusing to a normal area, such that the accuracy of a disparity value and the accuracy of a value domain in the optimized disparity map are ensured.
A vehicle control method and apparatus, a device, a product, and a storage medium. The method comprises: testing the working state of a vehicle (S301); determining, according to the working state of the vehicle, whether the vehicle needs to be dormant (S302); and when it is determined that the vehicle needs to be dormant, controlling, according to the working state of the vehicle, at least part of components configured on the vehicle to enter a dormant state (S303). The vehicle control apparatus comprises a detection module (401), a determination module (402), and a dormancy control module (403). The electronic device comprises a memory, a processor, and a computer program. A transport device comprises the vehicle control apparatus and the vehicle. The computer program product comprises a computer program or a computer instruction. The computer readable storage medium has the computer program stored. The present invention aims to prolong the working duration and improve the utilization rate of the electric quantity of the vehicle.
An image splicing processing method and apparatus, an electronic system and device, and a readable medium. The method comprises: obtaining a first overlapping image and a second overlapping image that respectively correspond to the overlapping regions of a first image to be spliced and a second image to be spliced (S202); determining a fused energy map corresponding to the first overlapping image and the second overlapping image (S204); determining a minimum energy path corresponding to the first map edge to the second map edge of the fused energy map (S206); according to the minimum energy path, determining a first mask corresponding to the first overlapping image and a second mask corresponding to the second overlapping image (S208); and splicing the first image and the second image according to the first mask and the second mask (S210). The method optimizes the region of a spliced seam to a certain extent, and improves the splicing effect of the image.
The embodiments of the present application provide an image processing method and apparatus. The method comprises: performing a conversion operation on each target module in an original model to obtain a processing model, wherein the original model is pre-trained, the target module comprises a plurality of branches, and the conversion operation comprises: for each branch in each target module, converting calculation parameters of the branch into parameters of a convolutional layer corresponding to the branch, the size of the convolutional layer corresponding to each branch being the same preset size, fusing the parameters of the convolutional layer corresponding to each branch to obtain convolutional layer parameters corresponding to the target module, and converting the target module into a convolutional layer corresponding to the target module having the convolutional layer parameters corresponding to the target module; and receiving an image to be processed, and processing the image by using the processing model.
A control method, a control device (60), a non-volatile readable storage medium, and a sorting system (100). The control method is applied to a control device (60) in a sorting system (100). The sorting system (100) comprises a goods transport mechanism and a goods pick-up mechanism that are communicatively connected to the control device (60). The goods transport mechanism comprises an unloading position. The control method comprises: obtaining a first image of the area in which the goods transport mechanism is located; determining, on the basis of the first image, location information of an item to be sorted when said item arrives at the unloading position; determining, on the basis of the location information, the transport speed of the goods transport mechanism, and item information of said item, transfer parameters of said item, the item information comprising the size and/or weight of said item, and the transfer parameters comprising at least one among transfer time, transfer angle, and transfer speed; and controlling the goods transport mechanism to, on the basis of the transfer parameters, transfer said item to the goods pick-up mechanism.
A management method for a sorting system, and a sorting system, a control device and a storage medium. The management method is applied to a control device of a sorting system. The sorting system comprises a camera and a sorting apparatus which are in communication connection with the control device. The sorting system, the control device and a storage medium comprise and store a computer-readable instruction for executing the management method. The management method comprises: acquiring an image, which is captured by a camera, of an area where a sorting apparatus is located; on the basis of the image of the area where the sorting apparatus is located, determining whether a sorting abnormality is present in the sorting process of the sorting apparatus; and if so, generating a sorting abnormality notification. By means of a sorting abnormality notification, a sorting abnormality situation can be handled in a timely manner, and therefore the sorting efficiency is improved.
B65G 43/00 - Control devices, e.g. for safety, warning or fault-correcting
B65G 15/00 - Conveyors having endless load-conveying surfaces, i.e. belts and like continuous members, to which tractive effort is transmitted by means other than endless driving elements of similar configuration
G06K 9/62 - Methods or arrangements for recognition using electronic means
37.
IMAGE REGISTRATION, FUSION AND SHIELDING DETECTION METHODS AND APPARATUSES, AND ELECTRONIC DEVICE
A method for detection of image registration includes: after acquiring registration data between a first image and a second image, performing grid segmentation to the first image and the second image (100), wherein the registration data include at least a registration-point coordinate and a registration-point displacement; calculating a homography matrix of each of grids within the first image with a corresponding grid within the second image (200); calculating a difference between a registration-point displacement of each of registration points and a homography-matrix displacement of a grid where the registration point is located as a displacement difference (300); and according to the displacement difference, determining an erroneous registration point (400), wherein a registration point whose displacement difference from the grid where the registration point is located satisfies a predetermined condition is determined as an erroneous registration point.
Provided are an image processing method, an image processing apparatus, an electronic system and a readable storage medium. The image processing method comprises: acquiring a plurality of initial images collected under different exposures, and a mask image corresponding to the plurality of initial images, wherein the mask image is used for indicating a foreground region in the plurality of initial images; fusing brightness information of the plurality of initial images to obtain initial brightness information; on the basis of the mask image, fusing the initial brightness information and brightness information of a reference frame image in the plurality of initial images to obtain final brightness information; and obtaining a target image on the basis of the final brightness information and color information of the reference frame image. According to the method, a mask image indicating a foreground region of a plurality of images with different exposures is obtained on the basis of the images, and then brightness information of a foreground region in a reference frame image and a non-foreground region in initial brightness information are fused by means of the mask image to ensure matching of a foreground region in a fused image and the foreground region in the reference frame image, thereby improving the visual effect of the foreground region in the fused image and also improving the overall effect of the fused image.
The present application relates to a warehousing system management method and device, a control terminal, a computer readable storage medium, and a warehousing system. When objects carried by a carrying mechanism not only comprise objects in a specified state but also comprise objects not in the specified state, by means of the warehousing system management method provided by embodiments of the present application, the objects in the specified state among the objects carried by the carrying mechanism can be moved out, and the objects not in the specified state are reserved, so that the effect that some specified objects on the carrying mechanism can be moved out can be achieved. The reserved objects not in the specified state can be returned to the carrying mechanism to continue to execute subsequent procedures according to an original process, so that the efficiency of replacement can be improved.
Provided are a method and apparatus for generating an HDR image, and an electronic device and a readable storage medium. The method comprises: acquiring an image to be subjected to extraction in a photographic scene; extracting a brightness feature of said image; determining, according to the brightness feature of said image, a light category of the corresponding photographic scene; acquiring, according to the light category of the photographic scene, a frame selection policy corresponding to the light category; and acquiring, according to the frame selection policy, a plurality of frame images of the photographic scene, and synthesizing an HDR image according to the frame images. Therefore, an image to be subjected to extraction is first acquired by means of a photographic scene, and a frame selection policy is determined by means of said image and is executed, so as to finally obtain a synthesized HDR image. The whole process avoids the case where a fixed frame selection policy cannot adapt to an extreme scenario, such that the extreme scenario is preventing from weakening a dynamic range and a visual effect of the HDR image, and the produced HDR image has a better dynamic range and a better visual effect.
The present application provides an image stitching method and apparatus, and a device and a storage medium. The method comprises: acquiring an original image to be processed, the original image consisting of a plurality of sub-images collected by a plurality of collectors; in the original image, dividing the image of a non-overexposed region into an overlapping region and a non-overlapping region, the overlapping region being a region where the pixel coordinates of the plurality of sub-images overlap each other in the original image; separately performing pixel processing on the overlapping region and the non-overlapping region to generate first pixel information of the overlapping region and second pixel information of the non-overlapping region; and performing image stitching on the overlapping region and the non-overlapping region according to the first pixel information and the second pixel information to generate a stitched image of the original image. According to the present application, the grid noise of the stitched image is reduced, so that the stitched image is more continuous and smoother.
A localization and mapping method, an apparatus, a robot (100), and a computer storage medium. The method comprises: performing linear feature extraction on obtained point cloud data at time t, and obtaining a corresponding plurality of line segments (S110); estimating the relative orientation of the robot (100) at time t relative to time t-1 (S120); performing, according to the relative orientation, line segment matching on the plurality of line segments corresponding to time t and a plurality of line segments obtained from performing linear feature extraction at time t-1, and obtaining a line to which each line segment belongs (S130); performing joint optimization on orientations of the robot (100) at each time and a two line parameterized representation format of all the lines, and obtaining line parameters of all the lines and orientations of the robot (100) at each time (S140). When representing a line, a two line parameterized representation format is used, and thus the amount of calculation can be reduced as much as possible, which can thereby improve calculation efficiency and shorten calculation time.
An image detection method and apparatus, a device and a medium. Said method comprises: obtaining multiple two-dimensional images collected at different times, wherein the multiple two-dimensional images comprise a reference object and a target object (S201); for each two-dimensional image among the multiple two-dimensional images, determining, according to position information of the reference object and the target object respectively on the two-dimensional image, a three-dimensional relative position of the target object relative to the reference object (S202); and determining an image detection result according to three-dimensional relative positions corresponding to the multiple two-dimensional images (S203). The image detection method can improve the accuracy of cross-line detection.
A target detection method and apparatus, and an electronic device. The method comprises: acquiring an image to undergo detection (S202); and inputting the image into a target detection model, and obtaining a target detection result (S204), the target detection result comprising the position and a score of a bounding box corresponding to a target. A process of training the above target detection model comprises: inputting an image sample in an image sample set into a student network model, and obtaining a student model detection result corresponding to each pixel in a first feature map of the image sample; acquiring, by means of a teacher network model, a teacher model detection result corresponding to the image sample; determining tag allocation information of the image sample according to the teacher model detection result, and calculating a loss function value of the student network model according to the tag allocation information and the student model detection result; and adjusting parameters of the student network model on the basis of the loss function value, and continuing to perform training until a target detection model is obtained. The method improves performance and training efficiency of a target detection model.
Embodiments of the present disclosure provide a method for controlling a working frequency of a TOF sensor, an apparatus, a device, and a computer-readable storage medium, and the method includes: inputting a target image frame into a preset face detection model for face detection to determine a face region in the target image frame; determining feature information of the face region, according to the face region and depth information of the target image frame acquired by the TOF sensor; and regulating and controlling the working frequency of the TOF sensor, according to the feature information and a preset working frequency of the TOF sensor. The method achieves dynamic regulation and control of the working frequency of the TOF sensor, and significantly improves the user experience.
G01S 7/4861 - Circuits for detection, sampling, integration or read-out
G06V 40/16 - Human faces, e.g. facial parts, sketches or expressions
G06T 7/521 - Depth or shape recovery from laser ranging, e.g. using interferometryDepth or shape recovery from the projection of structured light
G06T 7/73 - Determining position or orientation of objects or cameras using feature-based methods
G06T 7/55 - Depth or shape recovery from multiple images
G01S 7/48 - Details of systems according to groups , , of systems according to group
G01S 17/89 - Lidar systems, specially adapted for specific applications for mapping or imaging
G06Q 20/40 - Authorisation, e.g. identification of payer or payee, verification of customer or shop credentialsReview and approval of payers, e.g. check of credit lines or negative lists
46.
SEMANTIC SEGMENTATION MODEL TRAINING METHOD AND APPARATUS, AND IMAGE SEMANTIC SEGMENTATION METHOD AND APPARATUS
A semantic segmentation model training method, an image semantic segmentation method, a semantic segmentation model training apparatus, an image semantic segmentation apparatus, an electronic device, and a computer-readable storage medium. The semantic segmentation model training method comprises: acquiring a training set, the training set comprising multiple images as well as annotation information corresponding thereto, and the annotation information corresponding to any image comprising a segmentation box annotation and/or a semantic segmentation annotation (S11); extracting features of the images, and obtaining feature data of the images (S12); on the basis of the feature data, obtaining first segmentation frame information and first semantic segmentation information (S13); on the basis of the feature data, the first segmentation frame information, and the first semantic segmentation information, obtaining second segmentation frame information and second semantic segmentation information of the images (S14); determining a loss value on the basis of the second segmentation frame information and annotation information and/or on the basis of the second semantic segmentation information and annotation information (S15); and on the basis of the loss value, adjusting parameters of a semantic segmentation model (S16). In the semantic segmentation model training method, training data can be conveniently acquired, thus the training effect can be improved by means of massive, high-quality data.
A video frame interpolation method, a model training method, and a corresponding device. The video frame interpolation method comprises: acquiring a first video frame and a second video frame; utilizing, on the basis of the first video frame and the second video frame, a first neural network to calculate an optical flow between the first video frame and a first intermediate video frame and/or an optical flow between the second video frame and the first intermediate video frame; utilizing the optical flow between the first video frame and the first intermediate video frame to reverse map the first video frame to acquire a first mapped video frame, and/or, utilizing the optical flow between the second video frame and the first intermediate video frame to reverse map the second video frame to acquire a second mapped video frame; and determining the first intermediate video frame on the basis of the first mapped video frame and/or of the second mapped video frame. In the method, the accuracy of calculating the optical flow of an intermediate frame is increased; therefore, the image quality of the first intermediate video frame ultimately acquired is improved, and the efficiency of frame interpolation using the method is increased.
Provided by the present disclosure are an image processing method, an apparatus, an electronic device, and a computer-readable medium, the method comprising: obtaining an image to be processed, performing feature extraction on the image to be processed, and obtaining a multi-scale feature map; performing reinforcement processing on a portion corresponding to a salient object in the multi-scale feature map, and obtaining a multi-scale reinforced feature map; performing image restoration on the multi-scale reinforced feature map, and obtaining a salient object mask corresponding to the image to be processed. In the method of the present disclosure, after reinforcement processing is performed on the portion corresponding to the salient object in the multi-scale feature map, a feature map corresponding to the salient object in the obtained multi-scale reinforced feature map is more pronounced, and after lastly performing image restoration on the multi-scale reinforced feature map, the salient object mask segmented out and obtained is more accurate and more precise, and the technical problem of a sub-par segmentation effect when processing an image using a current salient object segmentation method is remedied.
The present application relates to the technical field of information processing, and provides a neural network model determination method and apparatus, and an electronic device, and a readable storage medium. The method comprises: obtaining an initial model parameter in a neural network model; determining a mathematical distribution corresponding to the initial model parameter; determining, according to the mathematical distribution, a quantization parameter corresponding to the initial model parameter; and performing quantization processing on the initial model parameter in the neural network model by using the quantization parameter, so as to obtain the updated neural network model. In the present application, quantization is performed on the initial model parameter in the neural network model by means of the determined mathematical distribution of the initial model parameter to quantify the initial model parameter within a reasonable range, thereby not only reducing the storage space of model parameters but also improving the computation rate and precision of the neural network model.
Embodiments of the present application provide an image processing method and apparatus, an electronic device, and a storage medium. The method comprises: determining all regions used for determining a region to be optimized, the all regions used for determining a region to be optimized comprising at least one of the following: at least one low-confidence region, at least one region corresponding to a weak texture region, and at least one region corresponding to a repeating texture region; determining, on the basis of all regions used for determining a region to be optimized, a region to be optimized in a disparity map; and optimizing disparity values of pixels in the region to be optimized to obtain an optimized disparity map. A region to be optimized in a disparity map is accurately determined, and disparity values of pixels in the region to be optimized in the disparity map are optimized to accurately optimize the disparity map, thereby obtaining an optimized disparity map having a better optimization effect.
The present application relates to an image jitter correction method and apparatus, and a computer device, a storage medium and a computer program product. The image jitter correction method comprises: acquiring target image frame data and target gyroscope data corresponding to the target image frame data; transforming the target gyroscope data, so as to obtain a target gyroscope mapping matrix corresponding to the target gyroscope data; inputting the target gyroscope mapping matrix into a preset matrix mapping network, and outputting a target image frame mapping matrix; and transforming the target image frame data by means of the target image frame mapping matrix, so as to obtain corrected image frame data, wherein the corrected image frame data is target image frame data which has been subjected to jitter correction. Therefore, the aims of enhancing a camera anti-jitter system by means of target gyroscope data and eliminating image jitter in a camera having an optical anti-jitter system are achieved.
Provided are an image processing method and apparatus, and an electronic device, a readable medium and a mobile terminal, which relate to the technical field of image processing. The method comprises: acquiring a first image collected by a first photographic apparatus, and determining a target ROI in the first image; calculating an ROI offset of the target ROI in the first image and a second image; acquiring the current magnification of the first image, and performing slicing processing on the ROI offset according to the current magnification, so as to obtain the current superimposition offset for performing image stereoscopic correction on the first image at the current moment; performing image stereoscopic correction on the first image on the basis of the current superimposition offset, so as to obtain a corrected first image; and if the current magnification is a preset switching magnification, switching a mobile terminal from the first photographic apparatus to a second photographic apparatus for display. By means of the present disclosure, the technical problem in the prior art of there being an obvious jump when performing imaging switching of a photographic apparatus is alleviated.
Embodiments of the present disclosure provide a filter array, an image sensor, a digital camera, a mobile phone, and a monitoring system. The filter array comprises a plurality of tiled minimal repeating units, and each minimal repeating unit comprises: red filter units configured to transmit red light; green filter units configured to transmit green light; blue filter units configured to transmit blue light; and at least one of yellow filter units configured to transmit yellow light and cyan filter units configured to transmit cyan light. By using a four-color (such as, red, blue, green, and cyan RGBC, or red, blue, green, and yellow RGBY) or five-color (red, blue, yellow, green, and cyan RGYBC) filter unit (such as, a light filter) to filter incident light, the present application can improve the light transmission efficiency so as to improve the brightness of an image obtained by means of parsing, and also can improve the display color effect of the obtained image by improving the color sensing capability of a pixel unit.
The present invention relates to an image splicing method and apparatus, a computer device, and a storage medium. The method comprises: obtaining images to be processed acquired from different visual angles, converting the images to be processed to be in a unified coordinate space, and obtaining converted images corresponding to the images to be processed; determining an overlapping region of the converted images; obtaining spatial domain optical flow information on the basis of pixel points of the converted images in the overlapping region, the spatial domain optical flow information being used for representing the degree of deviation of the pixel points in the overlapping region; and splicing the converted images according to the spatial domain optical flow information. By using the method, the image splicing quality can be improved.
Provided in the present disclosure are feature point tracking training and tracking methods, an apparatus, an electronic device, and a storage medium. The tracking training method comprises: obtaining two adjacent frames in a sample video, and using one frame as an initial frame and the other frame as a target frame; performing feature point detection on the initial frame to obtain feature point coordinates; using a Siamese feature extraction neural network to obtain a feature tensor corresponding to the initial frame and a feature tensor corresponding to the target frame; determining, from among the feature tensor corresponding to the initial frame, a feature vector corresponding to the feature point coordinates, performing local matching between the feature vector and the feature tensor corresponding to the target frame, and obtaining a matching score map; inputting the matching score map into a feature point tracking neural network, and obtaining predicted coordinates corresponding to the feature point coordinates; determining the loss value between the predicted coordinates and coordinates corresponding to the highest score in the matching score map; and according to the loss value, adjusting network parameters, and cyclically executing said steps until the loss value is converged. The present disclosure reduces the reliance on data annotation.
The present disclosure relates to a target detection method and apparatus, a computer device and a storage medium. The method comprises: performing feature extraction on an image to be detected to obtain n sets of first features of different scales, wherein n is an integer greater than 1; inputting the n sets of first features of different scales into a first feature fusion network, wherein the first feature fusion network comprises n feature fusion layers, and in an nth feature fusion layer, an nth set of first features is used as an nth set of second features; in an i-1th feature fusion layer, acquiring an ith set of second features and weight parameters corresponding to the ith set of second features, multiplying the ith set of second features by the weight parameters, and fusing the multiplied features with an i-1th set of first features to obtain an i-1th set of second features, until a first set of second features is obtained; and inputting n sets of second features to a detection network to obtain category information and location information of a target in the image. The method is advantageous in improving the accuracy of target detection.
The present application provides an action identification method and apparatus and an electronic device, and relates to the technical field of image processing. The method comprises: if a target object is detected from a video frame, obtaining a plurality of images containing the target object and optical flow images of the plurality of images; extracting an object trajectory feature of the target object from the plurality of images, and extracting an optical flow trajectory feature of the target object from the optical flow images of the plurality of images; and identifying the action type of the target object according to the object trajectory feature and the optical flow trajectory feature. According to the embodiments of the present invention, the action type of the target object is identified according to the trajectory information of the target object in the video frame and the optical flow information of the target object in the optical flow image of the image, and the temporal feature information and the spatial feature information of the target object are fused, so that the detection and recognition precision of the action type is effectively improved, the detection efficiency can be considered, and the overall detection performance is improved.
Provided are an indoor visual navigation method, apparatus, and system, and an electronic device. The method comprises: upload a collected indoor image to be positioned to a server by means of a mobile device; determine a camera pose by means of the server when the mobile device collects the indoor image, and send the camera pose corresponding to the indoor image to the mobile device; then the mobile device establishes an AR coordinate system aligned with a world coordinate system according to the camera pose corresponding to the indoor image, and plans a shortest route in a pre-imported indoor topological map according to destination information set by a user; and finally display, on an interface, the current preview image collected by the mobile device, and according to the AR coordinate system and the shortest route, superimpose, on the current preview image, a three-dimensional identifier for indicating a route travel direction. The present application can provide indoor navigation service for the user, and guide the user to conveniently arrive at a destination.
Provided are a multi-camera zoom control method and apparatus, and an electronic system and a storage medium. The method comprises: in the process of a first camera collecting an image, if the current set magnification input by a user is in a magnification transition zone, starting a second camera; acquiring a corresponding stereo correction matrix on the basis of calibration parameters of the first camera and the second camera; calculating a translation matrix on the basis of an acquired pixel position correlation between the same content regions of interest that correspond to a first zoomed image and a second zoomed image and in combination with the current set magnification, and then calculating a smooth transition transformation matrix in combination with the stereo correction matrix; and performing, by applying the smooth transition transformation matrix, affine transformation on an image output by the first camera, so as to obtain a display image of a device. According to the method, a stable translation quantity is obtained in a region of interest by means of template matching and on the basis of the stereo correction and alignment of an image, so as to realize the smooth transition of the region of interest, thereby improving the alignment stability of the region of interest.
The present application relates to the technical field of image processing. Provided are an image identification method and apparatus, and an electronic device and a computer-readable medium. The method comprises: during image identification, first extracting a feature of a target object image to be subjected to identification; then calculating a first feature distance between the feature of said target object image and a feature of a base library image, and obtaining a second feature distance between said target object image and the base library image according to the first feature distance and a target scaling parameter, wherein the target scaling parameter is related to the feature of said target object image; and determining an identification result of a target object in said target object image according to the second feature distance. As such, the feature distance from said target object image to the base library image is shortened by means of distance scaling transformation of the first feature distance under the target scaling parameter, so as to improve the passing rate of identification of an image photographed under dim light, top light or at a large angle, etc. without increasing an identification error rate.
A temperature correction method, device and system, which relate to the technical field of computers. Said method comprises: acquiring a target reference image frame pair containing a person, the target reference image frame pair including a target visible light image frame and a target thermal imaging image frame (S202); converting the target visible light image frame into a depth image frame (S204); inputting the depth image frame and the target thermal imaging image frame into an atmospheric inverse scattering model, the atmospheric inverse scattering model being a neural network model obtained by pre-fitting the relationship between depth and temperature correction value (S206); and performing, by means of the atmospheric inverse scattering model, temperature correction on the target thermal imaging image frame on the basis of the depth image frame, so as to obtain a target temperature (S208). The present invention can effectively improve the accuracy of personnel temperature measurement.
A distance determination method, apparatus and system, relating to the technical field of image processing. The distance determination method comprises: acquiring a primary visual image photographed by a primary camera and an original secondary visual image photographed by a secondary camera (S202); acquiring, by means of feature extraction and feature matching, an initial matching point pair between the primary visual image and the original secondary visual image (S204); correcting the original secondary visual image on the basis of the initial matching point pair and different constraint conditions sequentially, so as to obtain a target secondary visual image, the different constraint conditions comprising: a minimum rotation angle and a minimum parallax (S206); and determining a focusing distance according to the primary visual image and the target secondary visual image (S208). The focusing distance can be determined more accurately.
The embodiments of the present application provide a person re-identification method and apparatus, an electronic device, and a storage medium. Said method comprises: obtaining, by means of a feature extraction network, all sub-feature image sequences to be identified of an image of a person to be identified; on the basis of all said sub-feature image sequences of the image of the person to be identified and all reference sub-feature image sequences of the image of each reference person, calculating a degree of similarity between the image of the person to be identified and the image of each reference person; and on the basis of the similarity between the image of the person to be identified and the image of each reference person, generating a person re-identification result of the image of the person to be identified. In cases where the image of the person to be identified is a full-body image or a partial-body image of a person to be identified to which the image of the person to be identified belongs and/or in cases where the image of a reference person is a full-body image or a partial-body image of a reference person to which the image of a reference person belongs, the similarity between the image of the person to be identified and the image of the reference person can be calculated more accurately, and the person re-identification can be completed more accurately.
Provided are an image rectification method and device, and an electronic system. The image rectification method comprises: acquiring a first image and a second image of the same photographing object by means of a first photographing device and a second photographing device which are coaxially disposed; and rectifying the second image according to photographing parameters of the first photographing device and the second photographing device to obtain a second rectified image corresponding to the second image, such that the parallax between the second rectified image and the first image in a vertical direction or a horizontal direction is zero. In the method, by taking a first image as a reference, and by means of adjusting the photographing parameters of a first photographing device and a second photographing device, only a second image is rectified, thereby improving the operation efficiency of image rectification, and improving the accuracy and stability of an image rectification result.
Provided are an image processing method, apparatus and system, and an electronic device. The method comprises: acquiring an image to be processed, wherein said image is collected from a target display screen by means of a photographing apparatus; and by means of an image processing model, removing image fog from said image to obtain an image to be processed, from which fog is removed, wherein the image processing model is obtained by training pixel arrangement information of the target display screen. In the present application, by means of an image processing model, image fog in an image to be processed which is collected from a target display screen by means of a photographing apparatus can be removed, thereby improving the image quality, and also improving the user experience.
A temperature measuring device, comprising: a calibration source module (10); a camera module (20) used for collecting a real scene image and thermal sensing image of a target region in which a target (100) to be detected is located; a connection structure (30), one end of which is connected to the camera module (20), and the other end of which is used for the calibration source module (10) to be installed on same, so that the calibration source module (10) is located in the target region; and a main control module (40), which is communicatively connected to the camera module (20) for use in calculating temperature parameters of the target (100) according to the real scene image and the thermal sensing image. The connection structure (30) is used to install the calibration source module (10) on the camera module (20), so that the calibration source module (10) can be prevented from being too far away and causing deviations during temperature calibration, and the accuracy of temperature detection can be improved.
Disclosed in embodiments of the present disclosure are a behavior recognition method and device and a computer storage medium. The method comprises: acquiring an image to be identified, and inputting said image into a pre-trained neural network; obtaining outputs of the neural network, the outputs comprising a first output representing the probability that the behavior of a character is normal in the image, a second output representing the probability that the character is performing a first behavior, and a third output representing the probability that the character is performing a second behavior; and determining the behavior of the character according to the outputs. Hence, according to the embodiments of the present invention, the behavior of the character in the image may be determined by using the pre-trained neural network, and specifically, the neural network may extract rich visual features and may pay attention to the specific behavior of the character in the image, so that the accuracy of behavior identification is higher. Moreover, the behavior recognition method provided in the embodiments of the present invention may meet an instantaneity requirement, and may perform real-time operation, thereby meeting the recognition requirements of various different application fields. (FIG. 6)
The present disclosure relates to the technical field of machine learning, and provided are a neural network model quantification method and apparatus, an electronic device, and a computer-readable storage medium. The method comprises: acquiring a neural network model, wherein the neural network model comprises a convolutional layer, a normalization layer and a quantized activation layer, and output features of the quantized activation layer are integer-type features; converting the parameters of the convolutional layer into integer-type parameters; merging the normalization layer and the quantized activation layer to obtain a merged layer; and converting the parameters of the merged layer into integer-type parameters to obtain a quantized neural network model. The present disclosure reduces the computational complexity of the neural network model, so that the quantized neural network model may run faster on target devices that only support fixed-point or low-bit width operations.
The present application relates to a disparity map processing method and apparatus, a computer device and a storage medium. Said method comprises: down-sampling an initial disparity image to obtain at least one mask image; performing patch processing and pixel replacement processing on the mask image to obtain a processed mask image; determining an abnormal disparity value on the initial disparity map according to the processed mask image; and finally performing interpolation processing and filtering processing on the abnormal disparity value on the initial disparity map to obtain a target disparity image. In the described disparity map processing method, patch processing and pixel replacement processing is performed on the mask image, and such processing method can correct a disparity value in the initial disparity map which is calculated incorrectly due to factors such as repetitive textures or weak textures in the original photographed image and complexity of edge information in the initial disparity map. Therefore, the disparity map processing method provided in the present application can improve the quality of the disparity map.
The present invention relates to the technical field of display terminals, and provides a fingerprint recognition apparatus, a display panel, a device, and a fingerprint recognition method. The fingerprint recognition apparatus is provided below the display panel, and the display panel comprises a first fingerprint region and a second fingerprint region; the first fingerprint region comprises at least two pixel units, and the color of a first optical signal emitted by each of the pixel units is the same; the second fingerprint region comprises a plurality of subregions, and the colors of second optical signals emitted by at least two subregions are different; the fingerprint recognition apparatus comprises a first photosensitive sensing unit and a second photosensitive sensing unit; the first photosensitive sensing unit is configured to receive a first optical return signal; the second photosensitive sensing unit is configured to receive a second optical return signal. The present invention mitigates the technical problems of a complicated manufacturing process and high manufacturing cost of the existing living fingerprint recognition technology.
A temperature measurement method, a temperature measurement apparatus, an electronic device and a computer-readable storage medium. The method comprises: acquiring, by means of a visible light camera and a thermal imaging camera, an image frame pair including a target object, wherein the image frame pair comprises a visible light image and an infrared image that are collected at the same moment, and a black body is further arranged in an image collection area of the thermal imaging camera (S202); determining a measured temperature of the target object on the basis of the image frame pair (S204); carrying out black body detection on the infrared image to obtain a detection result of the black body, wherein the detection result comprises position information of the black body in the infrared image (S206); determining a measured temperature of the black body on the basis of the detection result of the black body and the infrared image (S208); and correcting the measured temperature of the target object according to the measured temperature of the black body and a preset temperature of the black body, and taking the corrected temperature as a temperature measurement result of the target object (S210). The accuracy of temperature measurement can be effectively improved.
An image processing method and apparatus, and a method and apparatus for training an image processing model, which relate to the technical field of image processing. The image processing method comprises: acquiring an original diffraction image (S302); inputting the original diffraction image into an image processing model (S304); and, by means of the image processing model, performing restoration processing on the original diffraction image, and obtaining a target standard image corresponding to the original diffraction image (S306). The described method can simplify a means of image restoration, effectively raise the quality of a restored target standard image, and improve the effect of a display screen displaying an image.
Provided are an image processing method and apparatus, a computation processing device, and a medium. The method comprises: acquiring an image to be processed (S101); and inputting the image to be processed into an image processing model, and on the basis of an output of a model to be processed, obtaining an image processing result of the image to be processed (S102), wherein the image processing model comprises a plurality of levels of feature processing sub-networks; each level comprises feature processing nodes of different depths; for each feature processing node of each level except the last level, an output feature map of the feature processing node is determined on the basis of a gating network comprised in the feature processing node, so as to form the dynamic image processing model; and the processing result of the image to be processed is obtained on the basis of an output feature map output by each feature processing node of the last level in the image processing model. By means of the above-mentioned method, a known network structure can be adaptively selected, which structure is suitable for images with different dimension distributions; and an actual operation computation amount is controlled, thereby reducing the computation amount.
Provided are a quick response code (200), a quick response code encoding method and apparatus, a quick response code decoding method and apparatus, and an electronic system. The quick response code (200) comprises a plurality of core encoding areas (210), each of the core encoding areas (210) comprising a specified information bit (220) and a detection information bit (230), wherein the specified information bit (220) comprises binary information corresponding to preset original information; and the detection information bit (230) comprises position information of the core encoding area (210). In the quick response code (200), a plurality of core encoding areas (210) are provided in the quick response code (200), thereby improving the anti-blocking capability of the quick response code (200).
G06K 19/06 - Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code
G06K 17/00 - Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups , e.g. automatic card files incorporating conveying and reading operations
75.
IMAGE SEGMENTATION METHOD AND APPARATUS, ELECTRONIC DEVICE AND READABLE STORAGE MEDIUM
An image segmentation method and apparatus, a device and a storage medium, which aim to improve the accuracy of image segmentation. The method comprises: obtaining image features of an image to be segmented (S11); performing an up-sampling operation on the image features to obtain a foreground feature map and a background feature map (S12); normalizing pixel values of pixel points of the foreground feature map and pixel values of pixel points of the background feature map to obtain a target region mask map and a background region mask map, wherein the pixel value of each pixel point in the target region mask map represents the probability of a pixel point that corresponds to the pixel point in the image belonging to a target region, and the pixel value of each pixel point in the background region mask map represents the probability of a pixel point that corresponds to the pixel point in the image belonging to a background region (S13); and according to the target region mask map and the background region mask map, segmenting the image (S14).
The present invention relates to the technical field of image processing, and provides a fingerprint image processing method and apparatus, an electronic device, and a computer readable medium. The method comprises: obtaining an original ground pattern image and a fingerprint image to be processed, wherein the original ground pattern image and the fingerprint image to be processed are images collected by the same fingerprint sensor, and the original ground pattern image and the fingerprint image to be processed are denoised images; in the original ground pattern image, cropping a ground pattern image corresponding to the fingerprint image to be processed, so as to obtain a target ground pattern image; and by combining the target ground pattern image and the fingerprint image to be processed, determining fingerprint information in the fingerprint image to be processed, and using the fingerprint information to determine a fingerprint image subsequent to removing a ground pattern. The present invention alleviates the technical problem that a traditional ground pattern removing technology has a poor ground pattern removing effect on a fingerprint image collected by a large-area under-screen fingerprint sensor.
An optimization method and apparatus for a computation graph, a computer device, and a storage medium. The method comprises: obtaining a computation graph of a computation network model; then inserting at least one check node in the computation graph; when running to each check node, obtaining the current performance margin by means of the current check node; then determining an optimization policy according to the current performance margin; and according to the optimization policy, optimizing a resource that needs to be consumed by a computation node after the current check node. According to the optimization method, by inserting the check nodes, obtaining the current performance margin of a computer device when the computer device runs to each check node, and according to the current performance margin, selecting the optimization policy meeting the actual running condition of the computer device to optimize the resource that needs to be consumed by the computation node, the resource use condition of each computation node during computation can be dynamically adjusted in the process that the computer device runs to the computation node, thereby improving the utilization rate of the resource on the computer device.
Component segmentation-based identification model training and vehicle re-identification methods and devices, relating to the technical field of image identification. Said method comprises: inputting an input image into a component segmentation network and a global feature extraction network of an identification model, to acquire a plurality of masking maps and a global feature map of the input image (S11); according to the plurality of masking maps and the global feature map, determining a plurality of local features corresponding to the plurality of masking maps on a one-to-one basis (S12); according to the plurality of local features of different input images, determining a viewing angle sensing distance between the different input images (S13); and constructing a loss function according to the viewing angle sensing distance, and updating parameters of the identification model according to the loss function, so as to complete the training of the identification model (S14). Performing ranking according to local features of different images and in combination with the Euclidean distances between global features can effectively improve the accuracy of vehicle re-identification.
A neural network joint compilation method, an apparatus and an electronic device. The method comprises: acquiring multiple neural network models to be compiled and computation graphs of said multiple neural network models (S302); determining model combination schemes for said multiple neural network models, and jointly compiling computation graphs of neural network models in each model combination scheme to obtain an optimized computation graph corresponding to each model combination scheme (S304); and determining among the optimized computation graphs a target optimized computation graph, and generating a target instruction according to the target optimized computation graph (S306). When the target instruction obtained by means of the compilation of the described method is run on target hardware, it can be ensured that the performance of the hardware is fully used, and the efficient running of the target instruction can further be ensured.
The present disclosure relates to the technical field of display terminals, and provided are an optical fingerprint identification apparatus, an optical fingerprint identification method, and a touch terminal. The optical fingerprint identification apparatus starts from the side close to an OLED display device and comprises, in order from top to bottom: a microlens layer, at least one diaphragm layer and a photosensitive sensor array; the microlens layer comprises a plurality of microlenses, the diaphragm layer comprises a plurality of diaphragm holes, and the photosensitive sensor array comprises a plurality of photosensitive sensing units; the microlenses and the diaphragm holes correspond one-to-one; each microlens corresponds to one photosensitive sensing unit; and a light signal from the direction of the OLED display device that is reflected by a finger passes through the microlenses, then passes through the diaphragm holes of the at least one diaphragm layer, and is received by the photosensitive sensing units. The optical fingerprint identification apparatus, optical fingerprint identification method and touch terminal provided by the present disclosure limit light receiving angles of reflected light signals by means of the diaphragm holes, which may prevent the crosstalk of light sensing by the photosensitive sensing units and improve the imaging effect of an optical imaging structure.
The present invention provides a night scene high dynamic range image generation method, device, and an electronic apparatus, the method comprises: acquiring multiple frames of normal exposure images and one frame of underexposure image; denoising the multiple frames of normal exposure images to obtain a normal exposure result image; denoising the underexposure image to obtain an underexposure result image; and fusing the normal exposure result image and the underexposure result image to obtain an HDR image. Thus, the night scene HDR image is generated only through one frame of underexposure image and multiple frames of normal exposure images, on one hand, the number of frames of the required composite image is reduced, so that the image acquisition time is shortened, and on the other hand, the underexposure image only needs to be simply denoised, denoising is not needed for multiple frames, and thus a large amount of denoising time is saved; thus, night scene high dynamic range image generation can be realized only through multiple frame denoising with low complexity, the whole generation process is simpler, and the user experience effect is rapidly improved.
09 - Scientific and electric apparatus and instruments
12 - Land, air and water vehicles; parts of land vehicles
42 - Scientific, technological and industrial services, research and design
Goods & Services
Agricultural machines; disintegrators; aerating pumps for
aquaria; incubators for eggs; hair clipping machines for
animals; salt machine; saws [machines]; calenders; paper
diapers making machines; industrial inkjet printing
machines; machines for the textile industry; tea
manufacturing machines; juice extractors, electric; tobacco
processing machines; sewing machines; machines for the
bicycle industry; brick making machines; batteries engine;
cord making machines; enamel making machines; light bulb
making machines; wrapping machines; briquette making
machines; washing machines [laundry]; embossing machines;
glass-working machines; fertilizer manufacturing machines;
electromechanical machines for chemical industry;
mineworking machines; cutters [machines]; machines for the
petrochemical industry; conveyors [machines]; metalworking
machines; steam engines; internal combustion engines
(non-land vehicles); wind power engine facility; pushpin
making machines; zipper making machine; industrial robots;
handling machines, automatic [manipulators]; electric wire
and cable manufacturing machines; handy hydraulic presses;
machines for the electronic industry; optical cold working
machines; gas separating installations; guns (spray) for
paint; alternators; filters for cleaning cooling air, for
engines; pumps [machines]; valves [parts of machines];
compressors [machines]; air cylinders [parts of machines];
axles for machines; bearings [parts of machines]; belts for
machines; welding machines, electric; machines and apparatus
for cleaning, electric; sorting machines for industry;
electroplating machines. Smartphone software applications, downloadable; humanoid
robots with artificial intelligence; computer software
platforms, recorded or downloadable; counters; apparatus to
check franking; automated teller machines [ATM]; ticket
dispensers; voting machines; face recognition apparatus;
check on work attendance tool; cameras [photography];
biochips; measuring instruments; stereoscopes; materials for
electricity mains [wires, cables]; chips [integrated
circuits]; heat regulating apparatus; lightning rods;
electrolysers; fire extinguishing apparatus; radiological
apparatus for industrial purposes; protective goggles; theft
prevention installations, electric; spectacles; solar
batteries; battery chargers; transparencies [photography];
egg-candlers; dog whistles; decorative magnets; portable
remote control stopper; refrigerator magnets; electrified
fences; electronic collars to train animals; sports
whistles; thermal imaging cameras; animal signalling rattles
for directing livestock; video monitors; computers; computer
hardware; computer peripheral devices; electronic
publications, downloadable; microcomputers; computer
programs, downloadable; computer programs, recorded;
computer operating programs, recorded; computer software,
recorded; computer software applications, downloadable;
weighing machines; measures; electronic notice boards;
mobile telephones; Internet messaging equipment; video
recorders; camcorders; monitoring apparatus, other than for
medical purposes; security surveillance robots. Electric vehicles; remote control vehicles, other than toys;
unmanned convey car; automatically guided [driverless]
material handling cars; pumps for bicycle tires; handtruck;
handling carts; sleighs [vehicles]; tires for vehicle
wheels; repair outfits for inner tubes; air vehicles;
electrical anti-theft installations for vehicles. Technological research; research and development of new
products for others; quality control; design and development
of multimedia products; surveying; chemistry services;
biological research; meteorological information; material
testing; industrial design; design of interior decor; dress
designing; computer software design; updating and
maintenance of computer software; platform as a service
[PaaS]; software as a service [SaaS]; cloud computing;
electronic data storage; monitoring of computer systems by
remote access; development of computer platforms; computer
system integration services; conversion of computer programs
and data, other than physical conversion; creating,
designing and maintaining web sites; authenticating works of
art; graphic arts design; cloud seeding; handwriting
analysis [graphology]; cartography services; business card
design; graphic design of promotional materials; rental of
meters for the recording of energy consumption; weighing of
goods for others; computer-aided graphic design; set design
for theatrical enterprises; updating of marine charts.
83.
IMAGE FUSION METHOD AND APPARATUS, ELECTRONIC DEVICE, AND READABLE STORAGE MEDIUM
An image fusion method and apparatus, an electronic device, and a readable storage medium. The method comprises: acquiring at least two raw images to be processed in the same scenario; using one image among the at least two raw images to be processed as a reference frame, using other images as supplemental frames, and separately determining the luminance relationship between each supplemental frame and the reference frame; for each supplemental frame, linearly adjusting the luminance of pixels in the supplemental frame on the basis of the luminance relationship so as to obtain an adjusted supplemental frame; and fusing each adjusted supplementary frame and the reference frame to obtain a fused image. Since each adjusted supplemental frame is adjusted on the basis of the luminance of the reference frame, the difference between the luminance of each adjusted supplemental frame and the luminance of the reference frame can be further reduced, which can effectively solve the problem in which the transition in luminance is unnatural in a finally obtained image due to a plurality of luminances existing in the image.
The present application relates to a method and device for image fusion, a computer device, and a storage medium. Comprising: acquiring multiple exposed imaged of different levels of exposure on the basis of a same target scene; acquiring a first exposed image fusion weight graph corresponding to each exposed image, where the first exposed image fusion weight graphs comprise fusion weights corresponding to pixels of the exposed images; acquiring the area of each overexposed region in each exposed image; with respect to each exposed image, utilizing the area of each overexposed region in the exposed image for smoothing filtering the first exposed image fusion weight graph corresponding to the exposed image to produce a second exposed image fusion weight graph corresponding to the exposed image; and performing image fusion processing with respect to the multiple exposed images on the basis of each second exposed image fusion weight graph to produce a fused image. This takes into consideration the characteristics of different overexposed regions and prevents a phenomenon of losing the details of a small overexposed region, thus allowing the acquired fused image to be more lifelike.
A cyclic edge-preserving smooth filtering method and apparatus, and an electronic device. The method comprises: acquiring a color image of a target scenario and an initial image to be filtered corresponding to the color image (S202); and taking the initial image to be filtered as a first image to be filtered (S204), and carrying out cyclic edge-preserving smooth filtering processing as follows: on the basis of the color image, carrying out edge-preserving smooth filtering processing on the current image to be filtered in order to obtain the present smooth-filtered image (S2041); on the basis of the present smooth-filtered image and the initial image to be filtered, updating the current image to be filtered (S2042); and on the basis of the updated image to be filtered, the present smooth-filtered image, and the initial image to be filtered, determining whether to perform the next cyclic edge-preserving smooth filtering processing (S2043). According to the method, it can be determined, according to the updated image to be filtered, the present smooth-filtered image and the initial image to be filtered, whether the cyclic edge-preserving smooth filtering processing can be finished in advance, without needing to reach a preset number of cycles, that is, the method can be self-adapted to the scenario of the color image, thereby reducing the time consumed for image processing.
09 - Scientific and electric apparatus and instruments
42 - Scientific, technological and industrial services, research and design
Goods & Services
Computer programs (downloadable software); computer
programs, recorded; computer software, recorded; computer
software applications, downloadable; data processing
apparatus; humanoid robots with artificial intelligence;
computers; computer hardware; computer peripheral devices;
electronic tags for goods; computer software platforms,
recorded or downloadable; wearable computers; face
recognition apparatus; automated teller machines [ATM];
counters; punched card machines for offices; check on work
attendance tool; weighing machines; measures; electronic
notice boards; Internet messaging equipment; mobile
telephones; monitoring apparatus, other than for medical
purposes; security surveillance robots; virtual reality
headsets; video recorders; camcorders; television apparatus;
self-motion advertising machine; dashboard cameras; wearable
video display monitors; cameras [photography]; thermal
imaging cameras; biochips; measuring instruments;
stereoscopes; cables, electric; integrated circuits; video
screens; transducer; goggles for sports; theft prevention
installations, electric; electronic access control systems
for interlocking doors; spectacles; solar batteries; battery
chargers; electrified fences. Industrial design; monitoring of computer systems by remote
access; technological research; research and development of
new products for others; updating of computer software;
computer software consultancy; rental of computer software;
platform as a service [PaaS]; computer software design;
software as a service [SaaS]; providing information relating
to computer technology and programming via a web site; cloud
computing; information technology [IT] consultancy;
electronic data storage; data security consultancy; quality
control; surveying.
87.
PRODUCT DEFECT DETECTION METHOD, DEVICE AND SYSTEM
The present application relates to the technical field of detection, and provided therein are a product defect detection method, device and system. The method comprises: acquiring multiple product images of a product to be detected, the capture angle and/or capture location being different for different product images; performing defect detection on the product images by using a neural network model trained in advance so as to obtain defect information of each product image, the defect information comprising defect location and/or defect type; determining an initial defect detection result on the basis of the defect information of each product image; verifying the initial defect detection result; and determining a final defect detection result on the basis of the verified initial defect detection result. The present application can effectively improve the accuracy of product defect detection.
Provided are image registration, fusion and shielding detection methods and apparatuses, and an electronic device. The image registration detection method comprises: obtaining registration data of a first image and a second image, and then carrying out grid segmentation on the first image and the second image (100), wherein the registration data at least comprises registration point coordinates and registration point displacement; calculating a homography matrix of each grid in the first image and a corresponding grid in the second image (200); calculating a difference value between registration point displacement of each registration point and homography matrix displacement of a grid to which the registration point belongs as a displacement difference; and determining an error registration point according to the displacement difference (400), wherein the registration point, the displacement difference between which and the grid to which the registration point belongs satisfies a preset condition, is the error registration point. Thereby, by detecting the registration points of the registered images, the error registration point in the registered images can be determined, such that registration accuracy can be further improved on the basis of original image registration.
09 - Scientific and electric apparatus and instruments
42 - Scientific, technological and industrial services, research and design
Goods & Services
Downloadable software computer programs for use in database management, use as a spread sheet, word processing; computer programs, recorded, for use in database management, use as a spread sheet, word processing; computer software, recorded, for use in database management, use as a spread sheet, word processing; computer software applications, downloadable, for mobile phones, portable media players, handheld computers, namely, software for use in database management, use in electronic storage of data; data processing apparatus; humanoid robots with artificial intelligence; computers; computer hardware; computer peripheral devices; electronic tags for goods; computer software platforms, recorded and downloadable, for application development, web hosting, database management; wearable computers in the nature of smart glasses, smartwatches; face recognition apparatus, namely, optical character recognition apparatus, iris recognition security devices; automated teller machines ATM; counters, namely, thread counters, Geiger counters; punched card machines for offices; check on work attendance tool in the nature of downloadable computer software for monitoring work attendance; weighing machines; measures, namely, measuring tapes, measuring rules; electronic notice boards; Internet messaging equipment in the nature of computers; mobile telephones; monitoring apparatus, other than for medical purposes, in the nature of wireless controllers to monitor the functioning of other electronic devices excluding gaming apparatus; security surveillance robots; virtual reality headsets; video recorders; camcorders; television apparatus for projecting purposes; self-motion advertising machine in the nature of motion-picture cameras for self-developing film; dashboard cameras; wearable video display monitors; photographic cameras; thermal imaging cameras; biochips for research or scientific purposes; measuring instruments, namely, instruments for measuring length, tool measuring instruments; stereoscopes; cables, electric; integrated circuits; video screens; transducer, namely, electrical transducers, electro-optic transducers; goggles for sports; theft prevention installations, electric, namely, security control panels and motion detectors, security surveillance robots; electronic access control systems for interlocking doors; spectacles; solar batteries; battery chargers; electrified fences Industrial design; monitoring of computer systems by remote access to ensure proper functioning, detect breakdowns, detect unauthorized access or data breach; technological research in the field of medical instruments, video surveillance equipment; research and development of new products for others; updating of computer software; computer software consultancy; rental of computer software; platform as a service (PaaS) featuring computer software platforms for use in database management, for use as a spreadsheet, for word processing; computer software design; software as a service (SaaS) services featuring software for use in database management, for use as a spreadsheet, for word processing; providing information relating to computer technology and computer programming via a web site; cloud computing featuring software for use in database management, for use as a spreadsheet, for word processing; information technology [IT] consultancy; electronic data storage; data security consultancy; surveying
90.
IMAGE PROCESSING METHOD, APPARATUS AND DEVICE AND COMPUTER-READABLE STORAGE MEDIUM
Provided are an image processing method, apparatus and device and a computer-readable storage medium. The method comprises: acquiring a plurality of key points of each region of an original image, and enabling each region to correspond to different body parts in a portrait; determining shape context histogram features of each region of the original image according to the plurality of key points of each region; determining a plurality of cost matrices for each region according to the shape context histogram features of each region of the original image and preset shape context histogram features of each material region image of a corresponding region; screening out, according to the plurality of cost matrices for each region, a material region image matching each region of the original image from each material region image; and splicing the material region images matching each region of the original image to obtain a target image corresponding to the original image. The method quickly achieves great similarity of the target image and the original image overall and in each region, and a beautification effect is generated.
09 - Scientific and electric apparatus and instruments
12 - Land, air and water vehicles; parts of land vehicles
42 - Scientific, technological and industrial services, research and design
Goods & Services
Agricultural machines, namely, cultivators, harvesters, disk harrows, seeders; disintegrators for chemical processing; aerating pumps for aquaria; incubators for eggs; hair clipping machines for animals; salt machine, namely, electric salt mills, electric salt grinder; saws machines; calenders being parts of machines; paper diapers manufacturing machines; industrial inkjet printing machines; machines for the textile industry, namely, textile making-up machines, machines for dyeing textiles; tea manufacturing machines, namely, tea processing machines, tea leaves sorting machines, machines for harvesting tea leaves; juice extractors, electric; tobacco processing machines; sewing machines; machines for the bicycle industry in the nature of bicycle assembling machines; brick making machines; batteries engine, namely, engines and motors for the generation of electricity; cord making machines; enamel making machines; light bulb making machines; wrapping machines; briquette making machines; washing machines for laundry; embossing machines; glass-working machines; fertilizer manufacturing machines, namely, blender machines for use with fertilizer in industrial and commercial applications; electromechanical machines for chemical industry, namely, electrochemical marking machines; mine working machines; cutters being machines, namely, tree stump cutters, trench cutters; machines for the petrochemical industry, namely, oil refining machines, oil-well pumping machines; conveyors being machines; metalworking machines, namely, perforating machines for metalworking, horning machines for metalworking; steam engines in the nature of steam pressure variation engines, steam marine engines not for land vehicles; internal combustion engines for non-land vehicles being boats; wind power engine in the nature of wind powered electricity generators for wind power facility; pushpin making machines; zipper making machine; industrial robots; handling machines in the nature of cargo handling machines, material handling machines being palletizers; automatic manipulators for forging presses and for forging machines; electric wire and cable manufacturing machines in the nature of wire forming machines, wire coiling machines; handy hydraulic presses for metalworking; machines for the electronic industry, namely, electronic pre-inked stamp making machines; optical cold working machines; gas separating installations, namely, centrifugal separators; alternators; filters for cleaning cooling air, for engines; pumps for machines; valves being parts of machines; compressors for machines; air cylinders as parts of machines in the nature of compressed air machine components being cylinders; axles for machines; bearings as parts of machines; belts for machines; welding machines, electric; machines and apparatus for cleaning, electric, in the nature of electric carpet cleaning machines, electric compressor cleaning machines; sorting machines for industry in the nature of rice grain sorting machines, sand sorting machines; electroplating machines Smartphone software applications, downloadable, for use in database management, use as a spread sheet, word processing; humanoid robots with artificial intelligence; computer software platforms, recorded and downloadable, for application development, web hosting, database management; counters, namely, thread counters, Geiger counters; automated teller machines ATM; ticket dispensers in the nature of video lottery ticket terminal dispensers, electronic parking lot ticket dispensers; voting machines; face recognition apparatus, namely, optical character recognition apparatus, iris recognition security devices; check on work attendance tool in the nature of downloadable computer software for monitoring work attendance; photographic cameras; biochips for research or scientific purposes; measuring instruments, namely, instruments for measuring length, tool measuring instruments; stereoscopes; materials for electricity mains, namely, electric wires, electric cables; chips being integrated circuits; heat regulating apparatus, in the nature of electric control devices for heating management; lightning rods; electrolysers; fire extinguishing apparatus; radiological apparatus for industrial purposes, in the nature of x-ray security scanners, x-ray full body security scanners; protective goggles, namely, dust protective goggles, personal protective equipment being safety goggles; theft prevention installations, electric, namely, security control panels and motion detectors, security surveillance robots; spectacles; solar batteries; battery chargers; photographic transparencies; egg-candlers; dog whistles; decorative magnets; portable remote control stopper in the nature of bark control devices for training dogs; refrigerator magnets; electrified fences; electronic collars to train animals; sports whistles; thermal imaging cameras; animal signaling rattles for directing livestock; video monitors; computers; computer hardware; computer peripheral devices; electronic publications, downloadable, in the nature of book, magazine, manual in the field of artificial intelligence; microcomputers; computer programs, downloadable, for use in database management, use as a spread sheet, word processing; computer programs, recorded, for use in database management, use as a spread sheet, word processing; computer operating programs, recorded; computer software, recorded, for use as a spreadsheet, word processing; computer software applications, downloadable, for mobile phones, portable media players, handheld computers, namely, software for use in database management, use in electronic storage of data; weighing machines; measures, namely, measuring tapes, measuring rules; electronic notice boards; mobile telephones; Internet messaging equipment in the nature of computers; video recorders; camcorders; monitoring apparatus, other than for medical purposes, in the nature of wireless controllers to monitor the functioning of other electronic devices excluding gaming apparatus; security surveillance robots Electric vehicles, namely, electrically powered motor vehicles, self-propelled electric vehicle; remote control land vehicles for transport, other than toys; unmanned convey car in the nature of a driverless car; automatically guided being driverless material handling cars; pumps for bicycle tires; hand trucks; handling carts; sleighs being vehicles for transport purposes; tires for vehicle wheels; repair outfits for inner tubes, namely, patches for repairing inner tubes, adhesive rubber patches for repairing inner tubes; air vehicles, namely, hot air balloons, air-cushion vehicles; electrical anti-theft installations for vehicles Technological research in the field of medical instruments, video surveillance equipment; research and development of new products for others; design and development of multimedia products; surveying; chemistry services, namely, chemistry consultation, research in the field of chemistry; biological research; meteorological information; material testing; industrial design; design of interior decor; dress designing; computer software design; updating and maintenance of computer software; platform as a service (PaaS) featuring computer software platforms for use in database management, for use as a spreadsheet, for word processing; software as a service (SaaS) services featuring software for use in database management, for use as a spreadsheet, for word processing; cloud computing featuring software for use in database management, for use as a spreadsheet, for word processing; electronic data storage; monitoring of computer systems by remote access to ensure proper functioning, detect breakdowns, detect unauthorized access or data breach; development of computer platforms; computer system integration services; conversion of computer programs and data, other than physical conversion; creating, designing and maintaining web sites; authenticating works of art; graphic arts design; cloud seeding; handwriting analysis being graphology; cartography services; business card design; graphic design of promotional materials; rental of meters for the recording of energy consumption; weighing of goods being cargo containers, aircraft for others; computer-aided graphic design; set design for theatrical enterprises; updating of marine charts
09 - Scientific and electric apparatus and instruments
42 - Scientific, technological and industrial services, research and design
Goods & Services
(1) Downloadable applications for mobile phones to provide access to the internet; humanoid robots with artificial intelligence; computer software platforms for social networking; revolution counters; automated teller machines [ATM]; ticket dispensers; voting machines; facial recognition software; punched card machines for offices; cameras [photography]; DNA chips; measuring tapes; stereoscopes; electric wires and cables; electronic chips for the manufacture of integrated circuits; heat regulators; lightning rods; electrolysis apparatus for producing hydrogen and oxygen from water; fire extinguishers; protective goggles for sports; anti-theft alarms, other than for vehicles; spectacles; solar batteries; battery chargers; transparencies for use in photography; egg-candlers; dog whistles; decorative magnets; electronic vehicle lock system including remote control; refrigerator magnets; electrified fences; electronic collars to train animals; sports whistles; thermal imaging cameras; animal signalling rattles for directing livestock; video monitors; computers; computer hardware; computer mouse; downloadable electronic publications in the nature of books; microcomputers; downloadable computer game programs; recorded computer operating programs; computer operating programs, recorded; computer application software for mobile phones for use in database management; downloadable computer game software; electronic weighing scales for personal use; measuring rulers; electronic notice boards; mobile telephones; Internet messaging software; video recorders; camcorders; video monitors; security surveillance robots. (1) Technological research in the field of computer hardware systems; research and development of new products for others; quality control testing services for industrial machinery; design and development of computers and programs for computers; technical surveying; research in the field of chemistry; biological therapeutic research services; meteorological information; material testing; industrial design; design of interior decor; dress designing; computer software design; updating and maintenance of computer software; platform as a service (PaaS) featuring computer software enabling users and systems to collaborate (view and change) on data across applications; software as a service ( saas ) services featuring non-downloadable software for encrypting data; cloud computing providing software for database management; electronic data storage for archiving business records; monitoring of computer systems by remote access; providing a platform as a service [PaaS] that allows computer software developers to collaboratively develop software via a website; computer system integration services; document data transfer and conversion from one media to another; creating, designing and maintaining web sites; authenticating works of art; graphic arts design; cloud seeding; handwriting analysis [graphology]; cartography services; business card design; graphic design of promotional materials; rental of meters for the recording of energy consumption; weighing of vehicles; computer-aided graphic design; set design for theatrical enterprises; updating of marine charts
Provided are an image processing method and apparatus. The image processing method comprises: acquiring an image to be processed; determining a target depth change type of a target background area of the image to be processed, wherein the target depth change type is used for representing a magnitude change rule of depth values of pixel points in the target background area; on the basis of the target depth change type, determining filtering parameters corresponding to each pixel point in the target background area, wherein for pixel points with different depth values, corresponding filtering parameters are different; and for each pixel point in the target background area, using a filtering parameter corresponding to the pixel point to filter the pixel point. By means of the technical solution provided in the embodiments of the present application, blurring of different degrees is performed on pixel points with different depth values in a background area, and the blurring of an image depth-of-field is more realistic.
The embodiment of the present application relates to the technical field of data processing. Provided are a living body testing method and apparatus, an electronic device and a readable storage medium. The living body testing method comprises: extracting multiple frames of video images from a video collected for an object to be tested; for each frame of a video image in the multiple frames of video images, determining, according to features of the frame of a video image, a first probability that the frame of a video image represents whether the object to be tested is a living body; and determining whether the object to be tested is a living body, according to first probabilities respectively corresponding to the multiple frames of video images. According to the living body testing method provided in the present application, the accuracy of testing a living body can be improved.
The present disclosure provides a method, an electronic device, and a medium for image fusion. The method includes calculating a fusion coefficient image M based on a first frame image I1 or based on both the first frame image I1 and a second frame image I2; calculating a first gradient D1 of the first frame image I1 and a second gradient D2 of the second frame image I2; calculating a preliminary fusion result J based on the image M, the first gradient D1 and the second gradient D2; and obtaining an output image I3 based on the image M, the first gradient D1, the second gradient D2 and the preliminary fusion result J, wherein brightness of the first frame image I1 is greater than brightness of the second frame image I2, and wherein the image M is used to mark fusion weights of pixels in the first frame image I1.
A method for controlling the working frequency of a TOF sensor, and an apparatus, a device, and a computer readable storage medium, the method comprising: inputting a target image frame into a preset face detection model to implement face detection, and determining a face area in the target image frame (S10); on the basis of the face area and depth information of the target image frame acquired by a TOF sensor, determining feature information of the face area (S20); and, on the basis of the feature information and a preset working frequency of the TOF sensor, performing adjustment and control of the working frequency of the TOF sensor (S30). The present method implements dynamic adjustment and control of the working frequency of the TOF sensor, and significantly improves the user experience.
Provided are a laser coarse registration method, device, mobile terminal and storage medium, the method comprises: determining the position information of obstacles around a robot; converting the position information of the obstacle into an obstacle image; acquiring map data of a pre-established probability map; converting the map data into a map image; carrying out matching processing on the obstacle image and the map image through a convolutional neural network to obtain a coarse registration position of the robot on the map image. The laser coarse registration method carries out convolution matching processing on the obstacle image and the map image obtained by the robot through the convolutional neural network, achieves the coarse registration of laser data collected by the robot and the probability map, the registration range is expanded, and the precision of coarse registration is improved.
09 - Scientific and electric apparatus and instruments
12 - Land, air and water vehicles; parts of land vehicles
42 - Scientific, technological and industrial services, research and design
Goods & Services
Agricultural machines; disintegrators; aerating pumps for
aquaria; saws [machines]; calenders; industrial inkjet
printing machines; machines for the textile industry; tea
manufacturing machines; juice extractors, electric; tobacco
processing machines; sewing machines; machines for the
bicycle industry; brick making machines; battery cable
rolling machines; cord making machines; enamel making
machinery; light bulb making machines; wrapping machines;
briquette making machines; washing machines [laundry];
embossing machines; glass-working machines; fertilizer
manufacturing machines; electromechanical machines for
chemical industry; mineworking machines; cutters [machines];
machines for the petrochemical industry; conveyors
[machines]; metalworking machines; steam engines; internal
combustion engines, other than for land vehicles;
wind-powered installations for generating electricity [wind
farms]; pushpin making machines; zipper making machines;
industrial robots; hydraulic machines; machines for the
electronic industry; optical cold working machines; gas
separating installations; air brushes for applying paint;
alternators; pumps [machines]; axles for machines; welding
machines, electric; machines and apparatus for cleaning,
electric; sorting machines for industry; electroplating
machines; incubators for eggs; hair clipping machines for
animals; handling machines, automatic [manipulators];
electric wire and cable manufacturing machines; filters for
cleaning cooling air, for engines; valves [parts of
machines]; compressors [machines]; air cylinders [parts of
machines]; bearings [parts of machines]; belts for machines. Computers; counters; time clocks [time recording devices];
weighing machines; measures; electronic notice boards;
cellular phones; video recorders; cameras [photography];
biochips; stereoscopes; materials for electricity mains
[wires, cables]; chips [integrated circuits]; heat
regulating apparatus; electrolysers; fire extinguishing
apparatus; radiological apparatus for industrial purposes;
protective goggles; theft prevention installations,
electric; eyeglasses; solar batteries; transparencies
[photography]; egg-candlers; computer hardware; computer
peripheral devices; electronic publications, downloadable;
microcomputers; computer programs, downloadable; computer
programs, recorded; computer operating programs, recorded;
computer software, recorded; computer software applications,
downloadable; smartphone software applications,
downloadable; humanoid robots with artificial intelligence;
computer software platforms, recorded or downloadable; dog
whistles; decorative magnets; electrified fences; electronic
collars to train animals; sports whistles; animal signalling
rattles for directing livestock; refrigerator magnets;
camcorders; monitoring apparatus, other than for medical
purposes; security surveillance robots; video surveillance
apparatus; battery chargers; lightning rods; thermal imaging
cameras; measuring instruments; apparatus to check franking;
automated teller machines [atm]; ticket dispensers; voting
machines; facial recognition software; computer software for
wireless network communications. Electric vehicles; remote control vehicles, other than toys;
automatically guided [driverless] material handling trucks;
pumps for bicycle tires; trolleys; sleighs [vehicles]; tires
for vehicle wheels; repair outfits for inner tubes; air
vehicles; electrical anti-theft installations for vehicles;
automatically guided [driverless] material handling
tractors; handling carts. Technological research; surveying; chemistry services;
biological research; meteorological information; material
testing; industrial design; design of interior decor; dress
designing; computer software design; authenticating works of
art; graphic arts design; research and development of new
products for others; quality control; updating and
maintenance of computer software; platform as a service
[PaaS]; software as a service [SaaS]; cloud computing;
electronic data storage; monitoring of computer systems by
remote access; design and development of multimedia
products; development of computer platforms; computer system
integration services; conversion of computer programs and
data, other than physical conversion; creating and
maintaining web sites; cloud seeding; handwriting analysis
[graphology]; cartography services; business card design;
graphic design of promotional materials; rental of meters
for the recording of energy consumption; weighing of goods
for others; computer-aided graphic design; theatrical set
design; updating of marine charts.
99.
Image processing method, image processing apparatus, and computer-readable storage medium
Provided are image processing method and apparatus for processing an input image using a convolutional neural network system, and a computer-readable storage medium. The convolutional neural network system that includes an input layer, an intermediate layer and an output layer, the image processing method includes: receiving the input image via the input layer; extracting image features of the input image via the intermediate layer; and outputting processing results for the input image via the output layer, wherein the intermediate layer includes at least one network block each of which includes a first convolutional layer, a first grouping rearrangement layer and a second convolutional layer that are cascaded.
Provided are neural network-based image processing method and apparatus, and a computer-readable storage medium. The image processing method includes: inputting an image into an optimized neural network; extracting, by the optimized neural network, image features of the image; and outputting the image features, wherein the optimized neural network is obtained by performing a first optimization process on at least one sub-layer in a pre-trained initial neural network, each sub-layer of the at least one sub-layer includes a convolutional layer, and the first optimization process comprises: for each sub-layer of the at least one sub-layer, determining one or more channels to be removed from a filter of the convolutional layer and removing said one or more channels, and optimizing parameters of remaining channels in the filter of the convolutional layer, so that error of output features of each optimized sub-layer is minimized.
