RU2789856C1 - Method, device and system for video processing - Google Patents

Abstract

FIELD: image processing techniques.

SUBSTANCE: invention relates to image processing techniques. The effect is achieved by obtaining a video to be processed, wherein the video to be processed is obtained by performing processing to remove features for one or more objects in the original video; receiving a processing request for restoring features for one or more objects to be processed, wherein one or more objects to be processed is one or more of the one or more objects for which the processing to remove the features has been performed; according to a processing request for restoring features for one or more objects to be processed, obtaining the feature image information corresponding to one or more objects to be processed, wherein the feature image information for one of the one or more objects to be processed contains information about pixel positions of all or part of the distinguishing features for one of one or more objects to be processed in the source video; according to the feature image information of the one or more objects to be processed, performing processing to restore the features of the one or more objects to be processed in the video to be processed.

EFFECT: securing the privacy of one or more specific objects and then allowing those objects to be displayed normally.

15 cl, 14 dwg, 2 tbl

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority under Chinese Patent Application No. 201911416668.3 entitled "Method, Apparatus and System for Video Processing" filed December 31, 2019, the contents of which are hereby incorporated by reference in their entirety.

FIELD OF TECHNOLOGY TO WHICH THE INVENTION RELATES

[0002] The disclosure of the present invention relates to the field of image processing, and more specifically, to methods, devices and systems for video processing.

BACKGROUND OF THE INVENTION

[0003] In the field of video surveillance, it is very important to protect the privacy of the filmed target.

[0004] Currently, in order to protect the confidentiality of the target object in the original results of video surveillance, processing is usually performed to hide ("occlusion") in relation to all targets contained in the original video. However, in some specific scenarios, it is required to perform privacy protection for one or more targets in the original surveillance results and not to perform privacy protection for one or more targets, but to display them normally; or for surveillance results where all targets are under privacy protection, you want to display some target images normally and keep privacy protection for other targets.

[0005] Thus, there is an urgent need to implement a video processing method capable of performing privacy protection or general display of specific targets as described above.

SUMMARY OF THE INVENTION

[0006] According to embodiments of the disclosure of the present invention, methods, devices and systems for video processing are provided to solve the problem of the inability to perform privacy protection for one or more of the specified objects, or the problem of the inability to restore privacy protection for one or more of the specified objects. The corresponding technical solutions are described below.

[0007] In accordance with the first aspect, a video processing method is provided, including:

[0008] obtaining a video to be processed, wherein the video to be processed is obtained by performing processing to remove features of one or more objects in the original video;

[0009] receiving a request for processing to restore features for one or more objects to be processed, wherein one or more objects to be processed is one or more of the one or more objects for which processing was performed to remove the features;

[0010] in accordance with a processing request for restoring features for one or more objects to be processed, obtaining feature image information corresponding to one or more objects to be processed, wherein the feature image information of one of the one or more objects to be processed contains information about the pixel positions of all or part of the distinguishing features of one of one or more objects to be processed in the source video;

[0011] according to the image information of the features of the one or more objects to be processed, performing processing to restore the features of the one or more objects to be processed in the video to be processed.

[0012] According to a second aspect, a video processing method is provided, including:

[0013] obtaining the original video;

[0014] receiving a processing request to remove features for one or more objects to be processed, wherein one or more objects to be processed is at least one object in the source video;

[0015] in accordance with a request for feature removal processing for one or more objects to be processed, obtaining feature image information corresponding to one or more objects to be processed, wherein the feature image information of one of the one or more objects to be processed comprises information about the pixel positions of all or part of the distinguishing features of one of one or more objects to be processed in the source video;

[0016] according to the feature image information of one or more objects to be processed, performing feature removal processing on the one or more objects to be processed in the original video.

[0017] According to a third aspect, a video processing apparatus is provided, comprising:

[0018] a first video acquisition module configured to obtain a video to be processed, the video to be processed is obtained by performing feature removal processing on one or more objects in the original video;

[0019] the first information obtaining module, configured to receive a processing request for restoring features for one or more objects to be processed, wherein one or more objects to be processed is one or more of the one or more objects for which processing has been performed for removal of distinguishing features; in accordance with a processing request for restoring features for one or more objects to be processed, obtaining feature image information corresponding to one or more objects to be processed, wherein the feature image information of one of the one or more objects to be processed contains position information pixels of all or part of the distinguishing features of one of one or more objects to be processed in the source video;

[0020] A feature recovery module configured to perform feature recovery processing for one or more objects to be processed in the video to be processed according to the feature image information of the one or more objects to be processed.

[0021] According to a fourth aspect, a video processing apparatus is provided, comprising:

[0022] a second video acquisition module, configured to acquire the original video;

[0023] a second information acquisition module, configured to receive a processing request for feature removal for one or more objects to be processed, wherein the one or more objects to be processed is at least one object in the source video; in accordance with a processing request for removing features for one or more objects to be processed, obtaining feature image information corresponding to one or more objects to be processed, wherein the feature image information of one of the one or more objects to be processed contains position information pixels of all or part of the distinguishing features of one of one or more objects to be processed in the source video;

[0024] A feature removal unit configured to perform feature removal processing for one or more objects to be processed in the source video according to the feature image information of the one or more objects to be processed.

[0025] According to a fifth aspect, a video processing system is provided, comprising a control platform and an image control terminal, wherein

[0026] the control platform is configured to obtain a video to be processed, wherein the video to be processed is obtained by performing feature removal processing on one or more objects in the original video; receiving a processing request to restore features for one or more objects to be processed, wherein one or more objects to be processed is one or more of the one or more objects for which the processing to remove the features has been performed; in accordance with a processing request for restoring features for one or more objects to be processed, transmitting to the image management terminal a request to obtain image information of the features for one or more objects to be processed, wherein the image information of the features of one of the one or more to be object processing contains information about the pixel positions of all or part of the distinguishing features of one of the one or more objects to be processed in the source video; obtaining feature image information for one or more objects to be processed transmitted by the image management terminal; according to the feature image information of the one or more objects to be processed, performing processing to restore the features of the one or more objects to be processed in the video to be processed;

[0027] the image management terminal is configured to receive a feature image information acquisition request for one or more objects to be processed; in accordance with the request for obtaining image information of distinctive features for one or more objects to be processed, determining image information of distinctive features of one or more objects to be processed; and transmitting feature image information of the one or more objects to be processed to the management platform.

[0028] According to a sixth aspect, a video processing system is provided, comprising a control platform and an image control terminal, wherein

[0029] the control platform is configured to receive the original video; receiving a processing request to remove features for one or more objects to be processed, wherein one or more objects to be processed is at least one object in the source video; in accordance with a processing request for removing features for one or more objects to be processed, transmitting to the image management terminal a request to obtain image information of the features for one or more objects to be processed, wherein the image information of the features of one of the one or more objects to be processed object processing contains information about the pixel positions of all or part of the distinguishing features of one of the one or more objects to be processed in the source video; obtaining feature image information of the one or more object to be processed transmitted by the image management terminal; according to the feature image information of the one or more objects to be processed, performing feature removal processing on the one or more objects to be processed in the source video;

[0030] the image management terminal is configured to receive a feature image information acquisition request for one or more objects to be processed; in accordance with the request for obtaining image information of distinctive features for one or more objects to be processed, determining image information of distinctive features of one or more objects to be processed; and transmitting feature image information of the one or more objects to be processed to the management platform.

[0031] According to a seventh aspect, a computer device is provided, comprising a processor and a memory that stores at least one instruction, which, when loaded into memory and executed by the processor, implements a video processing method corresponding to the first and second aspects described above.

[0032] According to an eighth aspect, a computer-readable storage medium is provided, comprising at least one instruction, which, when loaded into memory and executed by a processor, implements a video processing method corresponding to the first and second aspects described above.

[0033] According to a ninth aspect, there is provided a computer program product comprising an instruction that, when loaded into memory and executed by a processor in a computing device, implements a video processing method corresponding to the first and second aspects described above.

[0034] By implementing the technical solutions corresponding to the embodiments of the disclosure of the present invention, at least the following positive results can be achieved.

[0035] According to embodiments of the disclosure of the present invention, if processing is performed on one or more objects in the video to be processed, a processing request is first received to restore distinctive features for one or more objects to be processed, wherein one or more objects to be processed is one or more than an object that has been subjected to feature removal processing. Then, in accordance with the processing request for restoring the features for one or more objects to be processed, the image information of the features of the one or more objects to be processed can be obtained, wherein the image information of the features of one of the one or more objects to be processed contains information about the pixel positions of all or part of the distinguishing features of one of one or more objects to be processed in the source video. Further, when feature recovery processing is performed for one or more objects to be processed in a video to be processed according to the feature image information of one or more objects to be processed, it is not necessary to perform feature restoration for all objects, but only for one or more object to be processed, thereby achieving a higher level of privacy protection for other objects in the video to be processed.

BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Below, in order to more clearly describe the technical solutions corresponding to the embodiments of the disclosure of the present invention, a brief description of the accompanying drawings required to explain the embodiments is given. It is obvious that the drawings described below serve only as some examples of the disclosure of the present invention, and experts in this field of technology can, without significant creative effort, suggest other drawings based on these drawings.

[0037] FIG. 1 shows a flowchart for a video processing method according to an embodiment of the disclosure of the present invention.

[0038] FIG. 2 shows a flowchart for executing a video processing method according to an embodiment of the disclosure of the present invention.

[0039] FIG. 3 is a diagram of a video frame in accordance with an embodiment of the present invention.

[0040] FIG. 4 is a diagram of an image of an object in accordance with an embodiment of the disclosure of the present invention.

[0041] In FIG. 5 is a diagram of an image of an object in accordance with an embodiment of the disclosure of the present invention.

[0042] FIG. 6 is a block diagram of a video processing apparatus according to an embodiment of the present invention.

[0043] FIG. 7 is a block diagram of a video processing apparatus according to an embodiment of the present invention.

[0044] FIG. 8 is a block diagram of a control platform in accordance with an embodiment of the present invention.

[0045] FIG. 9 shows a flowchart of a video processing method according to an embodiment of the disclosure of the present invention.

[0046] FIG. 10 shows a flowchart for executing a video processing method according to an embodiment of the disclosure of the present invention.

[0047] FIG. 11 shows a flowchart of a video processing method according to an embodiment of the present invention.

[0048] FIG. 12 shows a flowchart of a video processing method according to an embodiment of the present invention.

[0049] FIG. 13 shows a flowchart of a video processing method according to an embodiment of the disclosure of the present invention.

[0050] FIG. 14 shows a flowchart of a video processing method according to an embodiment of the present invention.

DETAILED DESCRIPTION OF EMBODIMENTS

[0051] For a clearer presentation of the objectives, technical solutions and advantages of the present invention, the following embodiments of the disclosure of the present invention are described in more detail with reference to the accompanying drawings.

[0052] In the current state of the art, in order to protect the confidentiality of the target object, for all targets in the original results of video surveillance, hiding processing can be performed for image areas corresponding to all targets in video frames, and information about the positions of image areas that have undergone hiding processing , is saved. In addition, users who need to view video surveillance results can be assigned different levels of authorization, and only users with the highest authorization level can view the unhidden areas of video surveillance results. To avoid wasting memory space, the original surveillance video is not stored locally after hide is performed. In this case, if the user with the highest authorization level wants to view the surveillance video, the image areas subjected to the surveillance video hiding processing can be found based on the stored image area position information, and the hiding removal processing for the image areas is performed to obtaining the original video of the observation results. According to the above solution, if privacy protection is performed for a target in the surveillance results, hiding processing is required for all targets, and accordingly, in the processing for removing hiding, this process is also required to be performed for all targets.

[0053] According to an embodiment of the present invention, there is provided a video processing method capable of performing privacy protection or general display of specific targets. The method may be implemented by a management platform. The management platform may be a computer device hosted on the management side or a mobile device of the administrator. The administrator can view the surveillance video within his management area from the image management terminal through the management platform. The image management terminal may, for example, be a network video recorder (NVR, Network Video Recorder). The following describes embodiments in which an NVR is used as an image management terminal. In order to protect the privacy of an object, the administrator can perform processing to restore protection for one or more specified objects in the surveillance video using the video processing method implemented within the embodiments of the present invention so that the privacy of other objects in the surveillance video is not lost.

[0054] As shown in FIG. 1, the video processing method execution flow may include the following steps.

[0055] In step 101, the video to be processed is obtained.

[0056] Obtaining a video to be processed is performed by feature removal processing for one or more objects in the original video. The source video may be surveillance video captured by a security camera. The surveillance video may be a live surveillance video stream or a stored surveillance video over a period of time.

[0057] According to an embodiment, the control platform may receive video to be processed from the NVR. Depending on the different requirements, the acquisition of the video to be processed can be done in different ways.

[0058] For example, if an administrator wants to view the behavior of one or more specified objects in the surveillance video captured by the surveillance device for an area over a period of time, then the surveillance video of that area for the time period is the source video for the video to be processed. The administrator can send a request to receive the video to be processed to the NVR through the management platform, and the request to obtain the video to be processed may contain information about the time of shooting/capture and information about the shooting position. After obtaining the corresponding surveillance video, the NVR may first perform feature removal processing for one or more objects in the surveillance video to obtain a video to be processed. Here, the feature removal processing may include tiling one or more objects in the observation video, or performing adjustment of pixel values or pixel block positions for one or more objects in the observation video. Next, the NVR can transmit the received video to be processed to the management platform.

[0059] According to another example, the NVR may receive surveillance video captured by each real-time surveillance device as the source video, and perform feature removal processing on one or more objects in the source video to obtain the video to be processed, and then directly transfer the video to be processed to the management platform.

[0060] In another example, if the above source video is a video surveillance result captured in real time or a stored recorded video surveillance result, then the NVR can directly transmit the source video to the management platform, which performs the processing to remove features for one or more objects in the source video in order to obtain the video to be processed.

[0061] At step 102, a processing request is received to restore features for one or more objects to be processed, wherein one or more objects to be processed is one or more of one or more objects for which processing was performed to remove the features .

[0062] According to an embodiment, the management platform may receive a processing request to restore features for one or more objects to be processed in various ways, some of which are described below.

[0063] According to the first method, the control platform displays one or more images of one or more objects, wherein the one or more objects are located in the source video corresponding to the video to be processed for which the feature removal processing has been performed, and receives a processing request for restoring features for one or more objects to be processed, initiated by the operation of selecting at least one target image, wherein at least one target image is at least one of one or more images displayed on the management platform, and one or more images are images of one or more objects in the source video corresponding to the video to be processed, for which the processing to remove features has been performed.

[0064] If the management platform requests a video to be processed from the NVR, the NVR may directly transmit the source video to the management platform, and the management platform performs source video object identification to obtain an image of each object, and also performs feature removal processing for one or more source video objects in order to obtain the video to be processed. Next, the management platform may display images of one or more objects that have been subjected to feature removal processing.

[0065] Alternatively, if the management platform requests video to be processed from the NVR, the NVR may perform source video object identification to obtain a corresponding image of each object and pixel position information corresponding to one or more objects requiring processing to remove features . Next, the NVR transmits to the management platform a corresponding image of each object, pixel position information corresponding to one or more objects requiring de-distinguishing processing, and the original video. The control platform performs feature removal processing on one or more objects in the source video according to pixel position information corresponding to one or more objects requiring feature removal processing to obtain a video to be processed, and displays images of the one or more objects for which processing was performed to remove distinctive features.

[0066] Alternatively, if the management platform requests the video to be processed from the NVR, the NVR may perform object identification of the source video to obtain a corresponding image of each object, and perform processing to remove features of one or more objects in the source video to obtain the object to be processed. video. Next, the NVR transmits the corresponding image of each object and the video to be processed to the management platform. The management platform displays images of one or more objects that have been subjected to feature removal processing.

[0067] In addition, it should be noted that if the acquisition of the video to be processed is performed by processing the observation results captured in real time, then the images of objects displayed by the control platform may also be updated, for example, by adding an image of a new object or replacing an image of an existing one. object.

[0068] If it is desired to perform feature restoration for one or more objects to be processed, the administrator only needs to perform an operation of selecting one or more images for one or more objects to be processed in the displayed object images in order to initiate a processing request for the restoration of distinctive features. for one or more objects to be processed.

[0069] Thus, by displaying images to be selected by the administrator, objects appearing in the video to be processed can be more clearly displayed to the administrator, and thus the administrator can be aware of the objects in the video to be processed in real time, which helps perform feature recovery of the object of interest, view the position of the object of interest, and obtain information about what is happening with the object of interest.

[0070] According to the second method, one or more images for one or more objects to be processed are entered by the user, that is, a processing request for restoring features for one or more objects to be processed is transmitted to the management platform.

[0071] According to the second method, one or more images for one or more objects to be processed are entered by the administrator. In this case, the administrator cannot see objects in the current video to be processed, and one or more objects can be restored and displayed only if the administrator provides accurate images of one or more objects. Thus, better and better protection of objects in the current video can be provided.

[0072] According to the third method, the control platform may display the video to be processed, and the user may select at least one area of the target image corresponding to one or more objects to be processed in the video to be processed to initiate a request for processing to restore features for one or more object to be processed.

[0073] According to the third method, when a behavior of interest appears, the administrator can quickly view which of the objects behaves in the manner of interest by selecting an area of the image.

[0074] In step 103, in accordance with a processing request for restoring features for one or more objects to be processed, acquisition of image information of the features of the one or more objects to be processed is performed.

[0075] The image information of the features of one or more objects to be processed contains information about the pixel positions of all or part of the features of one of the one or more objects to be processed in the original video.

[0076] When implemented, the management platform may first determine the ID of the object to be processed according to the processing request for restoring the features for the object to be processed, and then obtain the feature image information corresponding to the object to be processed based on the ID of the object to be processed.

[0077] The following describes a method for obtaining feature image information corresponding to one or more objects to be processed.

[0078] The management platform or NVR may perform feature removal processing on one or more objects in the original video, and then the management platform or NVR may obtain feature image information corresponding to the one or more objects that are subjected to feature removal processing. Feature image information corresponding to the above object to be processed is included in feature image information corresponding to one or more objects that have been subjected to feature removal processing.

[0079] The following describes a method by which a management platform or an NVR performs feature removal processing for one or more objects in an original video.

[0080] First, using a pre-trained object detection model, object detection is performed for each video frame of the source video in order to obtain an image of the object in each video frame. The object image may be an image within a preset graphics area and contain an object, and the preset graphics area may be a rectangular area, a circle, or the like. If the feature image information is pixel position information of all object features in the original video , then the obtained information about the positions of the pixels of the image of the object in the video frame may be image information of the features of the object in the video frame. Information about the positions of the pixels of the image of the object in the video frame may be indicated by the coordinates of the specified positions of the preset graphics area corresponding to the image. For example, if the preset graphics area is rectangular, the pixel position information may be indicated by the pixel coordinates of the four vertices of the rectangular area or the pixel coordinates of the top left vertex, and the length and width of the rectangular area. In another example, if the preset graphics area is a circle, the pixel position information may be indicated by the coordinates of the center of the circle and the length of the radius of the circle.

[0081] Depending on the actual requirements, the detected object may be a biological sign of a living being, a license plate on a car, a driver's license, an identity card, and the like. In FIG. 3 is a block diagram illustrating the biometric image of each living being in one video frame.

[0082] Next, using the pre-trained target object matching model, object matching is performed for images of objects in different video frames to determine images belonging to the same object in different video frames, and then processing is performed to remove distinguishing features for one or more objects requiring processing to remove distinguishing features.

[0083] In the course of performing the above distinguishing feature removal processing, identical identifiers may be assigned to the same objects, and the assigned identifier is an object identifier. In addition, information about the correspondence between identifiers and object images can be stored. In addition, each video frame corresponds to a frame identifier, and the same object may correspond to feature image information in multiple video frames of the original video. Thus, a device (such as an NVR or a management platform) performing feature removal processing for one or more objects of the source video can store correspondence information between object IDs, video frame IDs, and feature image information, as shown below in table 1.

[0084] If the NVR performs feature removal processing for one or more objects in the original video, when the NVR transmits the video to be processed to the management platform, the NVR may send to the management platform the above correspondence information between the object IDs and the image information of the features. features corresponding to the video to be processed together with the video to be processed so that the management platform can keep the correspondence. Of course, in order to better protect the privacy of the objects in the video to be processed, the NVR can locally store compliance information without transmitting this information to the management platform. If the management platform performs feature removal processing for one or more objects in the source video, then the management platform may store the matching information locally.

[0085] Alternatively, the NVR may store the matching information locally and not transmit the information to the management platform, but only transmit the feature image information of one or more features to be processed to remove features and the original video. The control platform may then perform feature removal processing on the original video based on the feature image information of one or more objects requiring feature removal processing to obtain a video to be processed. In this way, the protection of said objects in the video to be processed can be improved, and the management platform can also perform protection recovery for said objects by using an appropriate feature restoration method in combination with a previously used feature removal processing method in a post-protection process for said objects.

[0086] In addition, according to the first and second methods, in which, in step 102, the management platform receives a processing request for restoring features for one or more objects to be processed, if the NVR transmits an image of the object to the management platform, the NVR may transmit to platform management object identifier together with the image of the object. When the object image is received by the control platform as a result of local processing of the source video, if the object image is stored, the object ID can be directly stored accordingly. According to the third example, the control platform can store the corresponding image area of each object in each video frame and, accordingly, the object ID, that is, store the video frame ID, the object ID, and the image area. In addition, the image area may be a rectangular area corresponding to an object obtained by performing object detection on the original video, and this area may be indicated by the top left vertex pixel position information and the length and width of the rectangular area, or the pixel position information of four vertices of the rectangular area.

[0087] Methods for determining one or more identifiers of one or more objects to be processed are described below, in combination with corresponding methods, in which the control platform receives a processing request to restore the distinguishing features for one or more objects to be processed.

[0088] In accordance with the above first method, after selecting at least one target image in the object images displayed by the manager, the management platform may locally request to obtain at least one object ID corresponding to at least one target image, as one or more identifiers of one or more objects to be processed.

[0089] In accordance with the above second method, after the administrator enters one or more images of one or more objects to be processed for each of the one or more objects to be processed, the management platform may perform matching on the stored image of each object processed to remove distinctive features in the original video, to determine the similarity between the image of the object to be processed and the stored image of each object, and obtain the ID of the object corresponding to the image having the maximum similarity as the ID of the object to be processed.

[0090] According to the above third method, after the administrator selects at least one target image area in a video frame, for each of the at least one target image area, the control platform performs an overlap comparison for the target image area and the stored image area of each object in video frame and determines an object ID corresponding to the most overlapped image region as the object ID of the object to be processed for the target image region.

[0091] Obtaining feature image information corresponding to one or more identifiers for one or more objects to be processed is described below.

[0092] If feature image information acquisition of one or more objects to be processed is performed and a correspondence between object IDs and feature image information is locally stored, the management platform may directly locally request the acquisition of feature image information corresponding to one or more objects to be processed, based on one or more identifiers of one or more objects to be processed. If the correspondence between the object IDs and the feature image information is stored in the NVR, a request for obtaining the feature image information containing one or more IDs of one or more objects to be processed can be transmitted to the NVR, and the NVR locally requests to obtain the feature image information corresponding to one or more objects to be processed, and then transmits the feature image information corresponding to the one or more objects to be processed to the management platform. If feature image information corresponding to one or more objects to be processed is requested from the NVR, the computing resources of the GPU (Graphics Processing Unit) in the NVR can be used to speed up the execution of the requests.

[0093] In step 104, in accordance with the image information of the features of one or more objects to be processed, processing is performed to restore the features of the one or more objects to be processed in the video to be processed.

[0094] In implementation, due to the fact that different processing methods for removing features are used, processing methods for restoring the features are also different.

[0095] In the case of using overlaying a tile image on an object image in the feature removal process: In the feature recovery process, the tile image corresponding to pixel position information in the feature image information corresponding to the object to be processed may be deleted.

[0096] In the case of using object pixel value adjustment or pixel block position adjustment during feature removal processing: the feature image information may also contain pixel values of the pixel block corresponding to the pixel position information, in addition to the pixel position information; and in the feature recovery process, the pixel values of the pixel block corresponding to the pixel position information in the feature image information in the video to be processed can be adjusted according to the pixel values included in the feature image information.

[0097] According to an exemplary implementation, feature restoration can only be performed for one or more object features, that is, in step S103, the object feature image information contains pixel position information of one or more object features in the original video. Accordingly, during the processing method for removing features for an object in the original video in step 103, after determining images belonging to the same object in different video frames, feature identification can be performed for each image belonging to the same object to determine the image information of the features. each image. Alternatively, for each image belonging to the same object, based on the previously stored information about the relative position of the features of the object in the object image, image information of the features of each object image can be determined. For example, for the biological features of an object such as a living being, the identifiable features may be biological sub-features included in the biological features, and for an object such as an identity card, the identifiable features may be a photograph, identification number, place of residence and etc. In FIG. 4 is a block diagram illustrating various features recognizable in an ID card image.

[0098] Accordingly, a single feature identifier may be assigned to one feature category, for example, the first biological sub-feature, the second biological sub-feature, and the third biological sub-feature of the biological feature may correspond to the same identifier. Identifiers of distinguishing features of one category of different objects may be the same, for example, the identifier of the first biological sub-feature included in the biological features of a living being 1 and the identifier of the first biological sub-feature included in the biological features of a living being 2 may be similar. In this case, the correspondence between the object identifiers and the feature image information can be completed in a manner as shown in Table 2 below.

[0099] If one or more identifiers of one or more objects to be processed are determined in step 103 described above, then one or more identifiers of the features of one or more objects to be processed may also be determined. In the same way, respectively, methods for determining one or more feature identifiers of one or more objects to be processed are described in combination with corresponding methods, during which the management platform receives a processing request to restore the features for one or more objects to be processed.

[00100] According to the above first method, after the administrator selects the target image in the displayed object images, the management platform also displays the feature images of the target image, and the target image can be selected by the administrator. The control platform may determine the object ID corresponding to the target image as the object ID to be processed, and determine the feature ID corresponding to the target feature image as the feature ID of the object to be processed.

[00101] In accordance with the second method presented above, after matching the image having the maximum similarity with the input image of the objects to be processed, the control platform also displays the feature images of this image, and the target feature image can be selected by the administrator. The control platform may determine the object ID corresponding to the image having the maximum similarity with the input image of the object to be processed as the object ID of the object to be processed, and determine the feature ID corresponding to the target feature image as the feature ID of the object to be processed.

[00102] According to the above third method, after the administrator selects a target image area, the management platform can determine the image area having the most overlap with the target image area, and then also display feature images of that image, and the administrator can select the target feature image. The control platform may determine the object ID corresponding to the image area with the largest overlap with the target image area as the object ID to be processed, and determine the feature ID corresponding to the target feature image as the feature ID of the object to be processed.

[00103] After receiving the feature image information of the object to be processed, the feature image information of the object to be processed can be determined in accordance with both the ID of the object to be processed and the ID of the feature of the object to be processed. If the correspondence information between the object IDs and the feature image information is stored locally, the management platform, based on the object ID and the feature ID of the object to be processed, can directly locally request to obtain the feature image information corresponding to both the object to be processed ID and the feature ID of the object to be processed. . If the correspondence between the object IDs and the feature image information is stored in the NVR, a request to obtain the feature image information containing the object ID and the feature ID of the object to be processed can be transmitted to the NVR, and the NVR locally requests to obtain the feature image information corresponding to both the ID. of the object to be processed and the feature ID of the object to be processed, and then transmits the feature image information to the management platform.

[00104] When performing feature recovery for the object to be processed, the management platform may perform recovery of only one or more features of the object to be processed. Thus, the recovery of distinguishing features can be carried out for a specific feature of the specified object, which makes it possible to eliminate the leakage of other information about the object. As shown in FIG. 5, the object to be processed is an identity card, the features to be restored are a photograph, name and place of residence, but the restoration of distinctive features does not affect the identification number of the person. In this way, the leakage of the personal identification number can be eliminated.

[00105] According to an exemplary implementation, the correspondence between object IDs and feature image information, the correspondence between object images and object IDs, and the correspondence between image regions and object IDs may be stored in the NVR for management purposes. If it is desired to obtain feature image information of the object to be processed, the control platform may send to the NVR a request to obtain feature image information of the object to be processed, wherein the feature image information acquisition request may only contain an object image or area information of the target image. In order to determine the ID of the object to be processed and the feature image information of the object to be processed, the NVR can take steps to implement these two determination processes and transmit to the management platform the set feature image information corresponding to the object to be processed.

[00106] According to an exemplary implementation, different permissions may be set for different levels of administrators. The processing authority may include the time period allowed for processing, the category of the object being processed, and so on. The time period allowed for processing may be, for example, "Monday to Wednesday" or "3:00 to 5:00", etc. The object category may be, for example, a living being, an ID person, car, etc.

[00107] Accordingly, the processing may include the following steps: obtaining the account ID of the target user that is currently registered with the management platform, and determining the target processing authority corresponding to the target user account ID. In the event that the target processing authority allows processing a request for processing to restore features for one or more objects to be processed, based on the image information of the features of one or more objects to be processed, processing is performed to restore the features for one or more objects to be processed. in the video to be processed.

[00108] When implemented, if the administrator wants to process the corresponding video, then the administrator can register in the management platform using his own target user account. Upon receipt by the management platform of a processing request to restore the distinguishing features for one or more objects to be processed, the management platform may first determine the appropriate target processing authority based on the registered target user account ID, and then, based on the stored relationship between the processing authority and the time period. , during which processing is allowed, determine the time period that allows processing corresponding to the target processing authority. The management platform can then determine whether the time of receipt of the above request for feature restoration processing is within a time period allowing processing corresponding to the target processing authority. If the time of receipt of the request is within the processing-allowable time period, it is determined that the processing request for feature restoration satisfies the target processing authority. Next, an operation of obtaining feature image information of one or more objects to be processed can be performed. In addition, based on the correspondence between the processing authorization and the category of the processing object, the category of the processing object corresponding to the target processing authorization can be determined. Next, it is determined whether the category of the object suitable for processing corresponds to the target processing authorization. If the processing object category matches the target processing authority, it is determined that the processing request for feature restoration satisfies the target processing authority. Next, an operation of obtaining feature image information of the object to be processed can be performed. Of course, the time period allowing processing and the category of the object suitable for processing corresponding to the target processing authority can be determined simultaneously. If both conditions are satisfied, then the feature image information acquisition operation of the object to be processed is allowed to be performed. By this method, a higher level of object privacy protection can be provided.

[00109] It should be noted that in this case, the video processing method according to the embodiments of the disclosure of the present invention can also be applied to process the same types of distinguishing features of a plurality of objects, for example, to process the first biological sub-features of an object A, an object B, and an object C. In addition, the method can also be applied to processing a variety of different types of objects, such as image processing of a living being and a vehicle.

[00110] According to embodiments of the disclosure of the present invention, if processing is performed on one or more objects in the video to be processed, a processing request is first received to restore distinctive features for one or more objects to be processed, wherein one or more objects to be processed is one or more than one or more objects for which processing has been performed to remove distinguishing features. Then, in accordance with a request for feature recovery processing for one or more objects to be processed, the feature image information corresponding to one or more objects to be processed can be obtained, wherein the feature image information corresponding to one of the one or more objects to be processed can be obtained. processing objects contains information about the pixel positions of all or part of the distinguishing features of one of one or more objects to be processed in the source video. Further, if feature recovery processing is performed for one or more objects to be processed in the video to be processed according to the feature image information of one or more objects to be processed, it is not required to perform feature restoration for all objects, but only for one or more of the object to be processed, thereby achieving a higher level of privacy protection for other objects in the video to be processed. By applying the method presented by the embodiments of the present invention, it is possible to rapidly perform feature restoration for one or more moving targets in a video.

[00111] As shown in FIG. 2, a video processing method is also provided within an embodiment of the present invention. The processing method algorithm may include the following steps.

[00112] In step 201, the source video is obtained.

[00113] The source video may be the original surveillance video to which no feature removal processing has been applied, or a video to which one or more of the objects to which feature removal processing has been applied. The source video may be an observation video captured in real time or a stored recorded observation video.

[00114] When implemented, the administrator may receive the source video from the NVR through the management platform while the NVR transmits the source video to the management platform.

[00115] At step 202, a processing request is received to remove features for one or more objects to be processed, wherein one or more objects to be processed are one or more objects in the source video.

[00116] According to an embodiment, the management platform may receive a processing request to remove features for one or more objects to be processed in various ways, some of which are described below.

[00117] According to the first method, the control platform displays images of objects and receives a processing request to remove features for one or more objects to be processed, triggered by an operation of selecting at least one target image, the objects being located in the source video.

[00118] The control platform may perform object identification on the source video to obtain images of objects in the source video, or the NVR performs object identification on the source video and sends the source video and object images in the source video to the control platform. The control platform receives an image of the objects in the source video and then displays the images of the objects present in the source video.

[00119] According to the second method, one or more images for one or more objects to be processed are entered by the user, that is, a processing request for removing features for one or more objects to be processed is transmitted to the management platform, and the management platform receives a processing request for feature removal for one or more objects to be processed, wherein the processing request for the removal of distinctive features for one or more objects to be processed contains one or more images for one or more objects to be processed.

[00120] According to the third method, the control platform displays the source video, and the user can select at least one area of the target image corresponding to one or more objects to be processed in the video to be processed to initiate a request for processing to remove features for one or more object to be processed. The control platform receives a processing request to remove features for one or more objects to be processed, triggered by the operation of selecting at least one area of the target image in the source video.

[00121] The specific implementation of step 202 is similar to the specific implementation of step 102 in the above embodiment, and will not be re-described further. The difference between the two steps is that, at step 202, the request received by the management platform is a request for feature removal processing, and the images of one or more objects displayed by the management platform are images of one or more objects for which no processing is performed for removal of distinguishing features in the original video.

[00122] In step 203, in accordance with a feature removal processing request for one or more objects to be processed, feature image information of the one or more objects to be processed is obtained, wherein the feature image information of one of the one or more objects to be processed contains information about the pixel positions of all or part of the distinguishing features of one of one or more objects to be processed in the source video.

[00123] When implemented, the control platform may first determine one or more identifiers of one or more objects to be processed in accordance with the request for processing to remove features for one or more objects to be processed, and then obtain the image information of the features of one or more objects to be processed. object based on one or more identifiers for one or more objects to be processed.

[00124] The following describes a method for obtaining feature image information corresponding to one or more objects to be processed.

[00125] First, the control platform or NVR performs object detection for each video frame in the source video in order to obtain an image of the object in each video frame. For example, using a pre-trained object detection model, object detection is performed for each video frame of the source video in order to obtain an image of the object in each video frame. The object image may be an image within a preset graphics area and contain an object, wherein the preset graphics area may be a rectangular area, a circle, or the like. If the feature image information is information about the pixel positions of all the features of an object in the original video, then the acquired pixel position information of the object image in the video frame may be image information of the object features in the video frame. Information about the positions of the pixels of the image of the object in the video frame may be indicated by the coordinates of the specified positions of the preset graphics area corresponding to the image. For example, if the preset graphics area is rectangular, the pixel position information may be indicated by the pixel coordinates of the four vertices of the rectangular area or the pixel coordinates of the top left vertex, and the length and width of the rectangular area. In another example, if the preset graphics area is limited to a circle, the pixel position information may be indicated by the coordinates of the center of the circle and the length of the radius of the circle.

[00126] Based on actual requirements, the detected object may be a living being, a license plate, a vehicle, an identity card, and the like. Next, using the pre-trained target matching model, object matching is performed for images of one or more objects in different video frames to determine the images belonging to the same object in different video frames.

[00127] In the course of performing the above processing, identical identifiers may be assigned to the same objects, and the assigned identifier is the identifier of the object. In addition, information about the correspondence between identifiers and images of objects can be stored. In addition, each video frame is associated with a frame ID, and an object may be associated with feature image information in each video frame. Thus, the same object can correspond to feature image information in multiple video frames of the original video. A device (such as an NRV or a control platform) performing object detection on an original video may store correspondence information between object IDs, video frame IDs, and feature image information.

[00128] If the NVR transmits the source video to the management platform, the NVR may transmit to the management platform the above correspondence information between object IDs and feature image information corresponding to the source video together with the source video so that the management platform can store the correspondence information. Of course, in order to better protect the privacy of the objects in the video, the NVR can store compliance information locally without transferring the information to the management platform. If the control platform performs object detection on the source video, then the control platform can store the matching information locally.

[00129] In addition, according to the first and second methods, in which, at step 202, the management platform receives a processing request to remove features for one or more objects to be processed, if the NVR transmits an image of the object to the management platform, then the NVR may transmit to platform management object identifier together with the image of the object. When an object image is received by the control platform as a result of local processing of the source video, if the object image is stored, the object identifier can be directly stored accordingly. According to the third method, the control platform can store the image area of each object in each video frame and, accordingly, the object ID, that is, store video frame IDs, object IDs, and image areas.

[00130] The following describes methods for determining one or more identifiers of one or more objects to be processed, in combination with corresponding methods, in which the control platform receives a processing request to remove features for one or more objects to be processed.

[00131] In accordance with the above first method, after selecting at least one target image in the object images displayed by the manager, the management platform may locally request that at least one object identifier corresponding to at least one target image be obtained as one or more identifiers of one or more objects to be processed.

[00132] In accordance with the above second method, after the administrator enters one or more images of one or more objects to be processed for each of the one or more objects to be processed, the management platform can perform a comparison with the stored image of each object in the original video in order to determine the similarity between the image of the object to be processed and the stored image of each object, and obtain an object ID corresponding to the image having the maximum similarity as the ID of the object to be processed.

[00133] In accordance with the above third method, after the administrator selects at least one target image area in the video frame for each of the at least one target image area, the control platform performs an overlap comparison for the target image area and the stored image area of each object in the video frame and determines an object ID corresponding to the region of the image with the largest overlap as the object ID of the object to be processed for the region of the target image.

[00134] Deriving feature image information corresponding to one or more objects to be processed based on one or more identifiers of one or more objects to be processed is described below.

[00135] When obtaining the feature image information of one or more objects to be processed, if the correspondence between the object IDs and the feature image information is stored locally, the management platform may directly request locally to obtain the feature image information corresponding to the one or more objects to be processed, on based on one or more identifiers of one or more objects to be processed. If the correspondence between the object IDs and the feature image information is stored in the NVR, a request for obtaining the feature image information containing one or more IDs of one or more objects to be processed can be transmitted to the NVR, and the NVR locally requests to obtain the feature image information corresponding to one or more objects to be processed, and then transmits the feature image information to the management platform.

[00136] The specific implementation of step 203 is similar to the specific implementation of step 103 in the above embodiment. The specific description of step 203 may refer to the relevant parts of step 103 and will not be re-described further.

[00137] In step 204, in accordance with the feature image information of the one or more objects to be processed, processing is performed to remove the features of the one or more objects to be processed in the original video.

[00138] When implemented, the control platform may adjust the position of the block of pixels corresponding to the position information of the pixels of the object to be processed in the source video, while the adjustment may be an arbitrary setting or a setting based on a predetermined rule, which is not limited to embodiments of the disclosure of the present invention. In addition, the adjustment may be performed in accordance with the pixel value in the pixel block according to the pixel position information of the object to be processed in the source video, and the adjustment may be an adjustment to an arbitrary pixel value or an adjustment to a preset pixel value, which is not limited to the disclosures. of the present invention. In addition, the mosaic image may be superimposed on the image corresponding to the pixel position information of the object to be processed in the original video.

[00139] In an exemplary implementation, feature removal may be performed on only one or more features of an object. That is, at step 203, the object feature image information contains information about the pixel position of one or more features in the original video. More specifically, reference may be made to the description of the relevant parts of the embodiment shown in FIG. 1.

[00140] More specifically, in step 203, after determining images belonging to the same object in different video frames, feature identification for each image belonging to the same object can be performed to determine image information of the features of each image. Alternatively, for each image belonging to the same object, based on the previously stored information about the relative position of the features of the object in the object image, image information of the features of each object image can be determined. For example, for a biological distinguishing feature of an object such as a living being, the identifiable features may be biological sub-features included in the biological distinguishing features, and for an object such as an identity card, the identifiable features may be a photograph, an identification number, a place residence, etc.

[00141] Accordingly, one feature identifier may be assigned to one feature category, for example, the first biological sub-feature, the second biological sub-feature, and the third biological sub-feature of the biological feature may correspond to the same identifier.

[00142] If one or more identifiers of one or more objects to be processed are determined in step 203 described above, then one or more identifiers of the features of one or more objects to be processed may also be determined. In the same way, respectively, methods for determining one or more feature identifiers of one or more objects to be processed are described in combination with corresponding methods, during which the management platform receives a processing request to remove the features for one or more objects to be processed.

[00143] According to the above first method, after the administrator selects the target image in the displayed object images, the management platform also displays the feature images of the target image, and the target image can be selected by the administrator. The control platform may determine the object ID corresponding to the target image as the object ID to be processed, and determine the feature ID corresponding to the target feature image as the feature ID of the object to be processed.

[00144] In accordance with the above second method, after matching the image having the maximum similarity with the input image of the object to be processed, the control platform also displays the feature images of this image, and the target feature image can be selected by the administrator. The control platform may determine the object ID corresponding to the image having the closest similarity to the input image of the object to be processed as the object to be processed ID, and determine the feature ID corresponding to the target feature image as the feature ID of the object to be processed.

[00145] According to the above third method, after the administrator selects a target image area, the management platform can determine the image area having the most overlap with the target image area, and then also display feature images of that image, and the administrator can select the target feature image. The control platform may determine the object ID corresponding to the image area with the largest overlap with the target image area as the object ID to be processed, and determine the feature ID corresponding to the target feature image as the feature ID of the object to be processed.

[00146] After receiving the feature image information of one or more objects to be processed, for each of the one or more objects to be processed, image information of the features of the object to be processed can be determined in accordance with both the ID of the object to be processed and the ID of the feature of the object to be processed. . If the correspondence information between the object IDs and the object feature image information is stored locally, the management platform can directly request locally to obtain the feature image information corresponding to both the object-to-process ID and the feature ID of the object-to-be-processed, based on the object-to-process ID, and identifier of the attribute of the object to be processed. If the correspondence between the object IDs and the feature image information is stored in the NVR, a request for obtaining the feature image information containing the object IDs and the feature IDs of one or more objects to be processed can be transmitted to the NVR, and for each of the one or more objects to be processed The NVR locally requests to obtain feature image information corresponding to both the object-to-be-processed ID and the feature-to-be-processed feature ID, and then transmits the feature image information to the management platform.

[00147] If the management platform performs feature removal on the object to be processed, then the feature removal may be performed on only one or more features of the object to be processed. Thus, when viewing the feature-deleted video, the user can approximately know what content is at the feature-deleted position, which helps him perform the feature-deleted video feature restoration processing based on actual requirements in particular.

[00148] According to an exemplary implementation, the correspondence between object IDs and feature image information, the correspondence between object images and object IDs, and the correspondence between image regions and object IDs may be stored in the NVR for management purposes. If it is desired to obtain feature image information of the object to be processed, the control platform may send a request to the NVR to obtain feature image information of the object to be processed, and the feature image information acquisition request may only contain the object image or target image region information. In order to determine the ID of the object to be processed and the feature image information of the object to be processed, the NVR can take steps to implement these two determination processes and transmit to the management platform the set feature image information corresponding to the object to be processed.

[00149] According to an exemplary implementation, different permissions may be set for different levels of administrators. The processing authority may include the time period allowed for processing, the category of the object being processed, and so on. Accordingly, the processing may include the following steps: obtaining the account ID of the target user currently registered with the management platform, and determining the target processing authority corresponding to the target user account ID. In the event that the target processing authority allows processing a processing request to remove features for one or more objects to be processed, based on the image information of the features of one or more objects to be processed, processing is performed to remove the features for one or more objects to be processed. in the original video. More specifically, reference may be made to the relevant parts of the description of the embodiment shown in FIG. 1, and further this information is not duplicated.

[00150] According to embodiments of the disclosure of the present invention, if processing is performed for one or more objects in the source video, first, a processing request is received to remove features for one or more objects to be processed. Then, according to the feature removal processing request for one or more objects to be processed, the feature image information corresponding to one of the one or more objects to be processed can be obtained, wherein the feature image information corresponding to the one or more objects to be processed can be obtained. processing objects contains information about the pixel positions of all or part of the distinguishing features of one of one or more objects to be processed in the source video. Further, if the feature removal processing is performed on one or more objects to be processed in the source video according to the image information of the features of the one or more objects to be processed, it is not required to perform the feature removal on all objects, but only on the one or more specified objects, thereby achieving a higher level of protection of the privacy of one or more of these objects. In addition, in combination with the requirements of the user, privacy protection can be carried out in particular for one or more of these objects. By applying the method presented by the embodiments of the disclosure of the present invention, one or more specified moving targets in a video can be quickly performed de-singing.

[00151] Based on the same technical idea, a video processing device is also proposed within the framework of the disclosure of the present invention. The device may be a control platform corresponding to the above embodiments. As shown in FIG. 6, the device includes a first video acquisition module 610, a first information acquisition module 620, and a feature recovery module 630.

[00152] The first video obtaining module 610 is configured to obtain a video to be processed, wherein the video to be processed is obtained by performing feature removal processing on one or more objects in the original video.

[00153] The first information obtaining module 620 is configured to receive a processing request for restoring features for one or more objects to be processed, wherein one of the one or more objects to be processed is one or more of the one or more objects for which the processing to remove distinctive features; in accordance with a processing request for restoring features for one or more objects to be processed, obtaining feature image information corresponding to one or more objects to be processed, wherein the feature image information of one of the one or more objects to be processed contains position information pixels of all or part of the distinguishing features of one of one or more objects to be processed in the source video.

[00154] The feature recovery unit 630 is configured to perform feature recovery processing for one or more objects to be processed in the video to be processed according to the feature image information of the one or more objects to be processed.

[00155] According to an exemplary implementation, the first information acquisition module 620 is configured to:

[00156] in accordance with the request for processing to restore distinctive features for one or more objects to be processed, determining one or more identifiers of one or more objects to be processed;

[00157] obtaining feature image information corresponding to one or more identifiers of one or more objects to be processed.

[00158] According to an exemplary implementation, the first information acquisition module 620 is configured to:

[00159] displaying images of objects that have been processed to remove distinctive features, while the objects are located in the original video;

[00160] receiving a processing request for restoring features for one or more objects to be processed, triggered by an operation of selecting at least one target image;

[00161] obtaining at least one identifier of at least one target image as one or more identifiers of one or more object to be processed.

[00162] According to an exemplary implementation, the first information acquisition module 620 is configured to:

[00163] receiving a processing request for restoring features for one or more objects to be processed, wherein the processing request for restoring distinctive features for one or more objects to be processed contains at least one image including one or more objects to be processed;

[00164] for each of the one or more objects to be processed in the stored images of the objects contained in the source video, determining the target image having the maximum similarity to the image containing the object to be processed, and obtaining the object identifier for the target image as the identifier of the object to be processed .

[00165] According to an exemplary implementation, the first information acquisition module 620 is configured to:

[00166] displaying the video to be processed;

[00167] receiving a processing request for restoring features for one or more objects to be processed triggered by an operation of selecting at least one area of a target image in the video to be processed;

[00168] according to the stored correspondence between object identifiers and image regions, obtaining at least one object identifier corresponding to at least one region of the target image as one or more identifiers of one or more objects to be processed.

[00169] According to an exemplary implementation, the first information acquisition module 620 is also configured to:

[00170] obtaining one or more identifiers of one or more objects to be processed;

[00171] wherein obtaining feature image information corresponding to one or more identifiers of one or more objects to be processed includes:

[00172] for each of the one or more objects to be processed, obtaining feature image information corresponding to both the object to be processed ID and the feature ID of the object to be processed.

[00173] According to an exemplary implementation, the first information acquisition module 620 is configured to:

[00174] according to the stored correspondence between the object IDs in the source video and the object feature image information, determining the feature image information corresponding to one or more IDs of one or more object to be processed; or

[00175] transmitting one or more identifiers of one or more objects to be processed to the image control terminal, so that the image control terminal determines the feature image information corresponding to one or more identifiers of one or more objects to be processed, according to the stored correspondence between the object identifiers in the original video and image information of distinctive features of objects; and receiving feature image information corresponding to one or more identifiers of one or more objects to be processed transmitted by the image control terminal.

[00176] According to an exemplary implementation, the first information acquisition module 620 is configured to:

[00177] in accordance with a processing request for restoring the features for one or more objects to be processed, transmitting a request for obtaining information of the features image for one or more objects to be processed to the image management terminal, wherein the request for obtaining the information of the features image used to instruct the image management terminal to transmit feature image information corresponding to one or more objects to be processed;

[00178] receiving feature image information of one or more objects to be processed.

[00179] In accordance with a possible implementation, the device also contains:

[00180] the first control module, configured to obtain the account ID of the target user currently registered in the control platform;

[00181] wherein the first information acquisition module 620 is configured to

[00182] determining the target processing authority corresponding to the target user's account ID;

[00183] The feature recovery module 630 is configured to:

[00184] in the event that the target processing authority allows processing a request for processing to restore features to one or more objects to be processed, to perform processing to restore the features to one or more objects to be processed in the video to be processed based on image information of the features signs of one or more objects to be processed.

[00185] The specific method of performing operations in each module of the apparatus according to the embodiments of the present invention has been described in detail within the scope of the method embodiments and is not further discussed in detail.

[00186] It should be noted that the video processing apparatus represented by the above embodiments is only illustrated using the above-described division into functional modules in the video processing process, but in actual applications, these functions may be distributed among different functional modules in order to perform operations based on requirements, that is, the internal structure of the device is divided into various modules to perform some or all of the functions described above. Moreover, the video processing apparatus represented by the above embodiments is based on the same idea as the video processing method, and its specific implementation may refer to embodiments of the method that will not be described again.

[00187] Based on the same technical idea, a video processing device is also proposed within the scope of the disclosure of the present invention. The device may be a control platform corresponding to the above embodiments. As shown in FIG. 7, the device includes a second video acquisition module 710, a second information acquisition module 720, and a feature removal module 730.

[00188] The second video acquisition module 710 is configured to obtain the original video;

[00189] The second information acquisition module 720 is configured to receive a processing request for feature removal for one or more objects to be processed, wherein the one or more objects to be processed is at least one object in the source video; in accordance with a processing request to remove features for one or more objects to be processed, obtain feature image information corresponding to one or more objects to be processed, wherein the feature image information of one of the one or more objects to be processed contains position information pixels of all or part of the distinguishing features of one of one or more objects to be processed in the source video;

[00190] The feature removing unit 730 is configured to perform feature removal processing for one or more objects to be processed in the source video according to the feature image information of the one or more objects to be processed.

[00191] According to an exemplary implementation, the second information acquisition module 720 is configured to:

[00192] in accordance with the request for processing to remove distinctive features for one or more objects to be processed, determine one or more identifiers of one or more objects to be processed;

[00193] obtaining feature image information corresponding to one or more identifiers of one or more objects to be processed.

[00194] According to an exemplary implementation, the second information acquisition module 720 is configured to:

[00195] displaying images of objects present in the source video; receiving a processing request for feature removal for one or more objects to be processed triggered by an operation of selecting at least one target image; obtaining at least one object identifier for at least one target image as one or more identifiers of one or more object to be processed.

[00196] According to an exemplary implementation, the second information acquisition module 720 is configured to:

[00197] receiving a processing request to remove features for one or more objects to be processed, wherein the processing request to remove distinctive features for one or more objects to be processed contains at least one image including one or more objects to be processed; in the stored images of the objects contained in the source video, determining the target image having the maximum similarity to the image containing one of the one or more objects to be processed, and obtaining the object identifier for the target image as the object identifier of one of the one or more objects to be processed.

[00198] According to an exemplary implementation, the second information acquisition module 720 is configured to:

[00199] displaying the original video; receiving a processing request for feature removal for one or more objects to be processed triggered by an operation of selecting at least one area of the target image in the source video; according to the stored correspondence between object identifiers and image regions, obtaining at least one object identifier corresponding to at least one region of the target image as one or more identifiers of one or more objects to be processed.

[00200] According to an exemplary implementation, the second information acquisition module 720 is also configured to:

[00201] obtaining one or more identifiers of one or more objects to be processed;

[00202] wherein obtaining feature image information corresponding to one or more identifiers of one or more objects to be processed includes:

[00203] for each of the one or more objects to be processed, obtaining feature image information corresponding to both the object to be processed ID and the feature ID of the object to be processed.

[00204] According to an exemplary implementation, the second information acquisition module 720 is also configured to:

[00205] according to the stored correspondence between the object identifiers in the source video and the object feature image information, determining the feature image information corresponding to one or more identifiers of one or more object to be processed; or

[00206] transmitting one or more identifiers of one or more objects to be processed to the image management terminal, such that the image management terminal determines feature image information corresponding to one or more identifiers of one or more objects to be processed according to the stored correspondence between object identifiers in the original video and image information of the distinguishing features of objects; and receiving feature image information corresponding to one or more identifiers of one or more objects to be processed transmitted by the image control terminal.

[00207] According to an exemplary implementation, the second information acquisition module 720 is also configured to:

[00208] Based on a processing request for removing features for one or more objects to be processed, transmitting to the image processing terminal a feature image information acquisition request, wherein the feature image information acquisition request is used to instruct the image management terminal to transmit feature image information corresponding to one or more objects to be processed;

[00209] receiving feature image information of one or more objects to be processed.

[00210] In accordance with a possible implementation, the device also contains:

[00211] a second management module, configured to obtain an account ID of a target user currently logging on to the management platform;

[00212] wherein the second information acquisition module 720 is configured to:

[00213] determining the target processing authority corresponding to the target user's account ID;

[00214] The feature remover 730 is configured to:

[00215] in the event that the target processing authority allows processing a request for feature removal processing for one or more objects to be processed, performing processing to remove features for one or more objects to be processed in the source video based on the image information of the features to be processed one or more objects to be processed.

[00216] The specific method of performing operations in each device module according to the embodiments of the present invention has been described in detail within the scope of the method embodiments and is not further discussed in detail.

[00217] It should be noted that the video processing apparatus represented by the above embodiments is only illustrated using the above-described division into functional modules in the video processing process, but in actual applications, these functions may be distributed among different functional modules in order to perform operations based on requirements, that is, the internal structure of the device is divided into various modules to perform some or all of the functions described above. Moreover, the video processing apparatus represented by the above embodiments is based on the same idea as the video processing method, and its specific implementation may refer to embodiments of the method that will not be described again.

[00218] FIG. 8 is a block diagram of a computing device in accordance with an embodiment of the present invention. The computer device may be a control platform according to the above embodiments. Control platforms 800 may vary greatly due to different configurations or operating characteristics. The control platform 800 may include one or more central processing units (CPU, Central Processing Unit) 801 and one or more memory 802, which stores at least one instruction, upon loading and execution of which the processor 801 implements any of the above video processing methods.

[00219] Based on the same technical idea, the present disclosure also provides a video processing system comprising a control platform and an image control terminal, wherein

[00220] the control platform is configured to obtain a video to be processed, wherein the video to be processed is obtained by performing feature removal processing on one or more objects in the original video; receiving a processing request to restore features for one or more objects to be processed, wherein one or more objects to be processed is one or more of the one or more objects for which the processing to remove the features has been performed; in accordance with a processing request for restoring features for one or more objects to be processed, transmitting to the image management terminal a request to obtain image information of the features for one or more objects to be processed, wherein the image information of the features of one of the one or more to be object processing contains information about the pixel positions of all or part of the distinguishing features of one of the one or more objects to be processed in the source video; obtaining feature image information of the one or more object to be processed transmitted by the image management terminal; according to the feature image information of the one or more objects to be processed, performing processing to restore the features of the one or more objects to be processed in the video to be processed;

[00221] the image management terminal is configured to receive a request to obtain feature image information for one or more objects to be processed; in accordance with the request for obtaining image information of distinctive features for one or more objects to be processed, determining image information of distinctive features of one or more objects to be processed; and transmitting the feature image information of the one or more objects to be processed to the management platform.

[00222] In an exemplary implementation, the control platform is configured to:

[00223] in accordance with the request to restore features for one or more objects to be processed, determine one or more identifiers of one or more objects to be processed, and send to the image management terminal a request to obtain image information of the features for one or more objects to be processed, wherein the request for obtaining information of the image of distinctive features for one or more objects to be processed contains one or more identifiers of one or more objects to be processed;

[00224] wherein the image management terminal is configured to determine feature image information corresponding to one or more identifiers of one or more objects to be processed, and according to the stored correspondence between the object identifiers in the source video and the object feature image information, to transmit to the management platform feature image information corresponding to one or more identifiers of one or more objects to be processed.

[00225] In an exemplary implementation, the control platform is configured to:

[00226] displaying images of objects present in the source video; receiving a processing request for restoring features for one or more objects to be processed, triggered by an operation of selecting at least one target image; and obtaining at least one object identifier of at least one target image as one or more identifiers of one or more object to be processed; or

[00227] receiving a processing request for restoring features for one or more objects to be processed, wherein the processing request for restoring distinctive features for one or more objects to be processed contains at least one image including one or more objects to be processed; for each of the one or more objects to be processed in the stored object images contained in the source video, determining a target image having the maximum similarity to the image containing the object to be processed, and obtaining an object identifier for the target image as the object to be processed identifier; or

[00228] displaying the video to be processed; receiving a processing request for restoring features for one or more objects to be processed, triggered by an operation of selecting at least one area of the target image in the source video; according to the stored correspondence between object identifiers and image regions, obtaining at least one object identifier corresponding to at least one region of the target image as one or more identifiers of one or more objects to be processed.

[00229] In an exemplary implementation, the control platform is also configured to:

[00230] obtaining a feature identifier of one or more objects to be processed;

[00231] wherein the request to obtain feature image information for one or more objects to be processed also includes one or more identifiers of the features of the one or more objects to be processed;

[00232] wherein the image management terminal is configured to:

[00233] determining, for each of the one or more objects to be processed, feature image information corresponding to both the object to be processed ID and the feature ID of the object to be processed, according to the stored correspondence between object IDs in the source video, object feature IDs, and image information of the distinguishing features of objects.

[00234] In combination with FIG. 9, the processing flow executed by the management platform and the image management terminal in the above system is described below.

[00235] In step 901, the control platform receives the video to be processed.

[00236] When implemented, the control platform may receive the video to be processed from the image control terminal, wherein the execution of obtaining the video to be processed is performed by performing the feature removal processing for one or more objects in the original video.

[00237] In step 902, the management platform receives a request for processing to restore features for one or more objects to be processed, wherein one or more objects to be processed is one or more of the one or more objects for which processing has been performed to remove distinctives. signs.

[00238] When implemented, the management platform may receive a processing request to restore features for one or more objects to be processed in various ways. For example, the management platform may display images of objects found in the source video and receive a processing request to restore features for one or more objects to be processed, triggered by an operation of selecting at least one target image. According to another example, the control platform may display a video to be processed and receive a processing request for restoring features for one or more objects to be processed triggered by an operation of selecting at least one area of a target image in the video to be processed. According to another example, the management platform may receive one or more images input by the user as a processing request for restoring features for one or more objects to be processed.

[00239] In step 903, the management platform sends to the image management terminal a request to obtain feature image information for one or more objects to be processed according to a processing request for recovering features for the one or more objects to be processed.

[00240] When implemented, the control platform may, after receiving a processing request for restoring features for one or more objects to be processed, directly send to the image management terminal a request to obtain image information of the features for one or more objects to be processed. In this case, due to different ways of receiving a processing request for restoring the features for one or more objects to be processed, the information included in the transmitted request for obtaining the feature image information may also differ. For example, if the management platform receives a processing request for restoring features for one or more objects to be processed triggered by an operation of selecting at least one target image, or receives a processing request for restoring features for one or more objects to be processed by obtaining one or more images entered by the user, the feature image information request may contain one or more images for one or more objects to be processed. For example, if the management platform receives a processing request for restoring features for one or more objects to be processed triggered by an operation of selecting at least one target image area in the source video, the request for obtaining the feature image information may comprise at least one target image area. Images. Preferably, the control platform may determine the identifier of one or more objects to be processed based on the received processing request for restoring the features of the object to be processed, and the request for obtaining the features image information transmitted by the control platform to the image management terminal may contain one or more identifiers of one or more object to be processed.

[00241] In step 904, the image management terminal determines the feature image information of the one or more object to be processed in accordance with the request to obtain the feature image information of the one or more object to be processed.

[00242] When implemented, the image management terminal may determine one or more identifiers of one or more objects to be processed in accordance with at least one image of one or more objects to be processed or at least one area of the target image included in the information request feature images for one or more objects to be processed, and then obtain image information of the features of one or more corresponding objects to be processed according to one or more identifiers for one or more objects to be processed, or obtain one or more identifiers of one or more objects to be processed included in the feature image information request for one or more objects to be processed, and to obtain the feature image information of one or more corresponding objects to be processed. objects based on one or more identifiers of one or more objects to be processed.

[00243] In step 905, the image management terminal sends to the management platform feature image information of one or more objects to be processed.

[00244] In step 906, the management platform performs feature recovery processing for one or more objects to be processed in the video to be processed according to the feature image information of the one or more objects to be processed.

[00245] When implemented, the control platform may perform feature restoration processing for one or more objects to be processed in the video to be processed by referring to the appropriate feature removal method according to the acquired image information of the features of the one or more objects to be processed. For example, a mosaic image superimposed on image areas corresponding to feature image information of one or more objects to be processed in the video to be processed may be removed. According to another example, the pixel value or pixel block sequence of image areas corresponding to image information of the features of one or more objects to be processed in the video to be processed may be changed to match the location in the original video.

[00246] In conjunction with FIG. 10, the processing flow performed by the management platform and the image management terminal in the above system will be described below.

[00247] In step 1001, the control platform receives the video to be processed.

[00248] In step 1002, the management platform receives a feature recovery processing request for one or more objects to be processed, wherein the one or more objects to be processed is one or more objects for which the feature removal processing has been performed, and determines one or more identifiers of one or more objects to be processed in accordance with the processing request to restore the distinguishing features for one or more objects to be processed.

[00249] When implemented, the management platform may receive a processing request to restore features for one or more objects to be processed in various ways. For example, the management platform may display images of objects found in the source video and receive a processing request to restore features for one or more objects to be processed, triggered by an operation of selecting at least one target image. According to another example, the management platform may receive one or more images input by the user as a processing request to restore features for one or more objects to be processed; and for each of the one or more objects to be processed, determine an image having the maximum similarity with the image entered by the user in the stored images as the target image; and then determining the ID of the object to be processed according to the above target image and the stored correspondence between the object images and the object IDs. In addition, the attribute identifier of the object to be processed can also be determined.

[00250] In step 1003, the management platform sends to the image management terminal a request to obtain feature image information for one or more objects to be processed, wherein the request to obtain feature image information for one or more objects to be processed contains one or more identifiers for one or more objects to be processed.

[00251] When implemented, the management platform sends to the image management terminal a feature image information acquisition request containing one or more identifiers of one or more objects to be processed. The feature image information request may also contain one or more identifiers of the above one or more object to be processed.

[00252] In step 1004, the image management terminal, according to the stored correspondence between the object IDs in the original video and the object feature image information, determines the feature image information corresponding to one or more IDs of one or more object to be processed.

[00253] When implemented, the image management terminal retrieves one or more object identifiers included in the feature image information request for one or more object to be processed. If the request to obtain feature image information for one or more objects to be processed also includes one or more feature identifiers of the one or more objects to be processed, the image management terminal may also extract one or more feature identifiers of the one or more objects to be processed. Then, for each of the one or more object to be processed, the image management terminal may determine feature image information corresponding to both the object ID and the feature ID of the object to be processed as the feature image information of the object to be processed.

[00254] In step 1005, the image management terminal sends to the management platform feature image information of one or more objects to be processed.

[00255] In step 1006, the management platform performs feature restoration processing for one or more objects to be processed in the video to be processed according to the feature image information of the one or more objects to be processed.

[00256] In combination with FIG. 11, the processing flow executed by the management platform and the image management terminal in the above system is described below.

[00257] In step 1101, the image management terminal transmits to the management platform the original video and image information of the features of the objects in the original video.

[00258] When implemented, the control platform may request the source video from the image control terminal, and the image control terminal may first perform processing operations such as object recognition, object matching, and feature recognition of the original video to obtain image information of one or more features of one or more objects in the original video, and then jointly transmit to the management platform the original video and image information of the features of one or more objects in the original video.

[00259] In step 1102, the control platform performs feature removal processing for one or more objects in the original video based on the feature image information of the one or more objects to obtain a video to be processed.

[00260] When implemented, the control platform may perform feature removal processing for one or more objects in the source video according to the received feature image information of the one or more objects to obtain a video to be processed. The feature removal processing in this case may be overlaying a tiled video, or adjusting pixel values or pixel block order.

[00261] It should be noted that the image management terminal may also perform feature removal processing for one or more objects in the original video and transmit the feature removed video directly to the management platform, and thus the management platform can receive the video to be processed.

[00262] In step 1103, the image management terminal transmits to the management platform one or more object identifiers and images of one or more objects in the source video.

[00263] In step 1104, the control platform displays images of the objects, and after performing an operation of selecting one or more images of one or more objects to be processed, determines one or more identifiers of one or more objects to be processed.

[00264] In step 1105, the management platform sends to the image management terminal a request to obtain feature image information for one or more objects to be processed, wherein the request to obtain feature image information for one or more objects to be processed contains one or more identifiers for one or more objects to be processed.

[00265] In step 1106, the image management terminal, according to the stored correspondence between the object identifier and the object feature image information in the source video, determines the feature image information corresponding to one or more identifiers of one or more object to be processed.

[00266] In step 1107, the image management terminal sends to the management platform feature image information of one or more objects to be processed.

[00267] In step 1108, the control platform performs feature restoration processing for one or more objects to be processed in the video to be processed according to the feature image information of the one or more objects to be processed.

[00268] It should be noted that the embodiments of the system and the processing algorithm corresponding to the system are based on the same idea as the above-described video processing method, and to describe a specific implementation, reference may be made to embodiments of this method, which will not be repeated hereinafter.

[00269] Based on the same technical concept, the present disclosure also provides a video processing system comprising a control platform and an image control terminal.

[00270] The control platform is configured to receive the source video, receive a processing request for feature removal for one or more objects to be processed, wherein the one or more objects to be processed is at least one object in the source video; in accordance with a processing request for removing features for one or more objects to be processed, transmitting to the image management terminal a request to obtain image information of the features for one or more objects to be processed, wherein the image information of the features of one of the one or more to be object processing contains information about the pixel positions of all or part of the distinguishing features of one of the one or more objects to be processed in the source video; obtaining feature image information of the one or more object to be processed transmitted by the image management terminal; according to the feature image information of the one or more objects to be processed, performing feature removal processing on the one or more objects to be processed in the source video.

[00271] The image management terminal is configured to receive a feature image information acquisition request for one or more objects to be processed; in accordance with the request for obtaining image information of distinctive features for one or more objects to be processed, determining image information of distinctive features of one or more objects to be processed; and transmitting the feature image information of the one or more objects to be processed to the management platform.

[00272] In an exemplary implementation, the control platform is configured to:

[00273] in accordance with the request to remove features for one or more objects to be processed, determine one or more identifiers of one or more objects to be processed, and send to the image management terminal a request to obtain image information of the features for one or more objects to be processed, wherein the request for obtaining information of the image of distinctive features for one or more objects to be processed contains one or more identifiers of one or more objects to be processed;

[00274] wherein the image management terminal is configured to determine feature image information corresponding to one or more IDs of one or more objects to be processed according to a stored correspondence between the object IDs in the original video and the feature image information of the objects; and transmitting to the management platform the feature image information corresponding to one or more identifiers of one or more objects to be processed.

[00275] In an exemplary implementation, the control platform is configured to:

[00276] displaying images of objects present in the original video; receiving a processing request for feature removal for one or more objects to be processed triggered by an operation of selecting at least one target image; obtaining at least one identifier of at least one target image as one or more identifiers of one or more object to be processed; or

[00277] receiving a processing request to remove features for one or more objects to be processed, wherein the processing request to remove distinctive features for one or more objects to be processed contains at least one image including one or more objects to be processed; for each of the one or more objects to be processed in the stored images of the objects contained in the source video, determining a target image having the maximum similarity to the image containing the object to be processed, and obtaining an object identifier for the target image as an identifier of the objects to be processed; or

[00278] displaying the original video; receiving a processing request for feature removal for one or more objects to be processed triggered by an operation of selecting at least one area of the target image in the source video; according to the stored correspondence between object identifiers and image regions, obtaining at least one object identifier corresponding to at least one region of the target image as one or more identifiers of one or more objects to be processed.

[00279] In an exemplary implementation, the control platform is configured to:

[00280] obtaining a feature identifier of one or more objects to be processed;

[00281] wherein the request to obtain feature image information for one or more objects to be processed also includes one or more identifiers of the features of the one or more objects to be processed;

[00282] wherein the image management terminal is configured to:

[00283] determining, for each of the one or more objects to be processed, feature image information corresponding to both the object to be processed ID and the feature ID of the object to be processed, according to the stored correspondence between object IDs in the source video, object feature IDs, and image information of the distinguishing features of objects.

[00284] In combination with FIG. 12, the processing flow executed by the management platform and the image management terminal in the above system is described below.

[00285] In step 1201, the control platform receives the original video.

[00286] When implemented, the control platform may perform the acquisition of the source video from the image control terminal, wherein the source video may be a video for which no feature removal processing has been performed, or a video containing one or more objects from which features have been removed.

[00287] At step 1202, the management platform receives a processing request to remove features for one or more objects to be processed.

[00288] When implemented, the management platform may receive a processing request to remove features for one or more objects to be processed in various ways. For example, the management platform may display images of objects found in the source video and receive a processing request to remove features for one or more objects to be processed, triggered by an operation of selecting at least one target image. According to another example, the control platform may display the source video and receive a feature removal processing request for one or more objects to be processed triggered by an operation of selecting at least one area of the target image in the source video. According to another example, the management platform may receive one or more images input by the user as a request for processing to remove features for one or more objects to be processed.

[00289] In step 1203, the management platform sends to the image management terminal a request to obtain feature image information for one or more objects to be processed according to a processing request for removing features for the one or more objects to be processed.

[00290] When implemented, the control platform may, upon receiving a processing request for feature removal for one or more objects to be processed, directly send to the image management terminal a request to obtain image information of the features image for one or more objects to be processed based on the request for processing to remove distinctive features. In this case, due to different methods of obtaining a feature removal processing request for one or more objects to be processed, the information included in the transmitted feature image information acquisition request may also differ. For example, if the management platform receives a processing request to remove features for one or more objects to be processed, triggered by an operation of selecting at least one target image, or a processing request to remove features for one or more objects to be processed, generated by obtaining one or more images entered by the user, the feature image information request may contain one or more images for one or more objects to be processed. According to another example, if the control platform receives a processing request for feature removal for one or more objects to be processed triggered by an operation of selecting at least one target image area in the source video, the feature image information acquisition request may comprise at least one target image area. Preferably, the control platform may determine the identifier of the one or more objects to be processed based on the received processing request to remove the features of the one or more objects to be processed, and the feature image information acquisition request transmitted by the control platform to the image control terminal may comprise one or more identifiers of one or more objects to be processed.

[00291] In step 1204, the image management terminal determines the feature image information of the one or more object to be processed in accordance with the request to obtain the feature image information of the one or more object to be processed.

[00292] When implemented, the image management terminal may determine one or more identifiers of one or more objects to be processed in accordance with the image of one or more objects to be processed or at least one area of the target image included in the feature image information request for one or more objects to be processed, and then obtain feature image information of one or more corresponding objects to be processed according to one or more identifiers for one or more objects to be processed, or obtain one or more identifiers of one or more objects to be processed included in the request to obtain feature image information for one or more objects to be processed, and to obtain feature image information of one or more corresponding objects to be processed based on one or more identifiers of one or more objects to be processed.

[00293] In step 1205, the image management terminal sends to the management platform feature image information of one or more objects to be processed.

[00294] In step 1206, the control platform performs feature removal processing for one or more objects to be processed in the source video according to the feature image information of the one or more objects to be processed.

[00295] When implemented, the control platform may perform feature removal processing for one or more objects to be processed in the source video according to the feature image information of the one or more objects to be processed. For example, a tiled image is superimposed on an image area corresponding to feature image information of one or more objects to be processed in the original video. According to another example, a pixel value or a pixel block sequence of an image area corresponding to feature image information of one or more objects to be processed in the original video can be adjusted.

[00296] In combination with FIG. 13, the processing flow executed by the management platform and the image management terminal in the above system is described below.

[00297] In step 1301, the control platform receives the original video.

[00298] In step 1302, the control platform receives a processing request to remove features for one or more objects to be processed, and determines one or more identifiers of one or more objects to be processed in accordance with the processing request to remove the features of one or more objects to be processed. object.

[00299] When implemented, the control platform may receive a processing request to remove features for one or more objects to be processed in various ways. For example, the management platform may display images of objects found in the source video and receive a processing request to remove features for one or more objects to be processed, triggered by an operation of selecting at least one target image. According to another example, the management platform may receive user input images as a request for processing to remove features for one or more objects to be processed; and for each of the one or more objects to be processed, determine an image having the maximum similarity with the image entered by the user in the stored images as the target image; and then determining the ID of the object to be processed according to the above target image and the stored correspondence between the object image and the object ID. In addition, the attribute identifier of the object to be processed can also be determined.

[00300] In step 1303, the management platform sends to the image management terminal a request to obtain feature image information for one or more objects to be processed, wherein the request to obtain feature image information for one or more objects to be processed contains one or more object identifiers one or more objects to be processed.

[00301] When implemented, the management platform sends to the image management terminal a feature image information acquisition request containing one or more identifiers of one or more objects to be processed, wherein the feature image information acquisition request may also contain one or more feature identifiers one or more objects to be processed.

[00302] In step 1304, the image management terminal, according to the stored correspondence between the object IDs in the source video and the object feature image information, determines the feature image information corresponding to one or more IDs of one or more object to be processed.

[00303] The image management terminal allocates one or more object identifiers included in the feature image information request to one or more object to be processed. If the identifiers of one or more features of one or more objects to be processed are included in the request, the image management terminal may also extract one or more of the feature identifiers of the one or more objects to be processed at the same time. Then, for each of the one or more object to be processed, the image management terminal may determine feature image information corresponding to both the object ID and the feature ID of the object to be processed as the feature image information of the object to be processed.

[00304] In step 1305, the image management terminal sends to the management platform feature image information of one or more objects to be processed.

[00305] In step 1306, the control platform performs feature removal processing for one or more objects to be processed in the source video according to the feature image information of the one or more objects to be processed.

[00306] In combination with FIG. 14, the processing flow executed by the management platform and the image management terminal in the above system is described below.

[00307] In step 1401, the image management terminal transmits the original video to the management platform.

[00308] In step 1402, the image management terminal sends the object ID and the image of the object in the source video to the management platform.

[00309] When implemented, the image management terminal can perform processing such as object recognition, object matching, and feature recognition of the original video, assign the same object ID to the same object in the original video, and obtain an image of the object, and then transmit the object image and the corresponding object identifier in the management platform.

[00310] In step 1403, the management platform displays object images and, after performing an operation of selecting one or more images of one or more objects to be processed, determines one or more identifiers of one or more objects to be processed.

[00311] In step 1404, the management platform sends to the image management terminal a request to obtain feature image information for one or more objects to be processed, wherein the request to obtain feature image information for one or more objects to be processed contains one or more identifiers for one or more objects to be processed.

[00312] In step 1405, the image management terminal, according to the stored correspondence between object IDs and object feature image information in the source video, determines feature image information corresponding to one or more IDs of one or more object to be processed.

[00313] In step 1406, the image management terminal sends to the management platform feature image information of one or more objects to be processed.

[00314] In step 1407, the control platform performs feature removal processing for one or more objects to be processed in the source video according to the feature image information of the one or more objects to be processed.

[00315] It should be noted that the embodiments of the system and the processing algorithms corresponding to the system are based on the same idea as the embodiments of the above video processing method, and for the description of a specific implementation, reference may be made to the embodiments of the video processing method, which are not repeated. are described.

[00316] The illustrated embodiment also provides a computer-readable storage medium comprising at least one instruction that, when loaded into memory and executed by a processor, implements a video processing method according to any of the embodiments described above. For example, the computer-readable storage medium can be Read-Only Memory (ROM), Random-Access Memory (RAM), CD-ROM, magnetic tape, floppy disk, optical storage device, and the like.

[00317] The above embodiment also provides a computer program product comprising an instruction that, when loaded into memory and executed by a processor in a computing device, implements a video processing method according to any of the embodiments described above.

[00318] Those skilled in the art can appreciate that all or part of the steps in the embodiments described above can be implemented using hardware, or by transferring appropriate instructions to the hardware using a program that can be stored on a computer-readable storage medium. This storage medium may be a Read Only Memory (ROM), a magnetic disk or a CD, or the like.

[00319] The above is a description of only some variants of the disclosure of the present invention, without limiting its scope. Any changes, equivalent replacements and improvements that do not contradict the essence and principles of the present invention, should be included in the scope of the invention.

Claims (54)

1. Video processing method applicable to the management platform, including: obtaining a video to be processed, which is obtained by performing a feature removal processing on one or more objects in the original video; receiving a processing request for restoring features for one or more objects to be processed, wherein one or more objects to be processed is one or more of the one or more objects for which the processing to remove the features has been performed; in accordance with a processing request for restoring features for one or more objects to be processed, obtaining feature image information corresponding to one or more objects to be processed, wherein the feature image information of one of the one or more objects to be processed contains position information pixels of all or part of the distinguishing features of one of one or more objects to be processed in the source video; And according to the feature image information of the one or more objects to be processed, performing processing to restore the features of the one or more objects to be processed in the video to be processed. 2. The method according to claim 1, wherein obtaining feature image information corresponding to one or more objects to be processed according to a processing request for restoring the features to one or more objects to be processed includes: in accordance with the request for processing to restore the distinguishing features for one or more objects to be processed, determining one or more identifiers of one or more objects to be processed and obtaining feature image information corresponding to one or more identifiers of one or more objects to be processed. 3. The method of claim 2, wherein receiving a processing request to restore features for one or more objects to be processed and determining one or more identifiers of one or more objects to be processed in accordance with the request for processing to restore the features for one or more more objects to be processed include: displaying one or more images of one or more objects, wherein one or more objects in the source video is processed to remove distinctive features; receiving a processing request for restoring features for one or more objects to be processed, triggered by an operation of selecting at least one target image; obtaining at least one object identifier of at least one target image as one or more identifiers of one or more objects to be processed; or receiving a processing request for restoring features for one or more objects to be processed, wherein the processing request for restoring distinctive features for one or more objects to be processed contains at least one image including one or more objects to be processed; for each of the one or more objects to be processed in the stored object images contained in the source video, determining a target image having a maximum similarity to the image containing the object to be processed, and obtaining an object identifier for this target image as an object to be processed identifier; or displaying the video to be processed; receiving a processing request for restoring features for one or more objects to be processed, triggered by an operation of selecting at least one area of the target image in the video to be processed; and, according to the stored correspondence between the object identifiers and image regions, defining at least one object identifier corresponding to at least one region of the target image as one or more identifiers of one or more objects to be processed. 4. The method of claim 2, wherein, before obtaining the feature image information corresponding to one or more identifiers of one or more objects to be processed, the method also includes: obtaining one or more identifiers of one or more objects to be processed; while obtaining image information of distinctive features corresponding to one or more identifiers of one or more objects to be processed includes: for each of the one or more objects to be processed, obtaining feature image information corresponding to both the object to be processed ID and the feature ID of the object to be processed. 5. The method according to claim 2 or 3, wherein obtaining the feature image information corresponding to one or more identifiers of one or more objects to be processed includes: according to the stored correspondence between the object IDs in the source video and the object feature image information, determining the feature image information corresponding to the one or more IDs of the one or more objects to be processed; or transmitting one or more IDs of one or more objects to be processed to the image management terminal, so that the image management terminal determines feature image information corresponding to one or more IDs of one or more objects to be processed according to the stored correspondence between the object IDs in the original video and the information images of distinctive features of objects; and receiving feature image information corresponding to one or more identifiers of one or more objects to be processed transmitted by the image control terminal. 6. The method according to p. 1, including also: obtaining an account ID of the target user currently logging into the management platform; wherein, before performing the feature recovery processing for one or more objects to be processed in the video to be processed according to the feature image information of the one or more objects to be processed, the method also includes: determining the target processing authority corresponding to the target user's account ID; wherein performing feature recovery processing for one or more objects to be processed in the video to be processed according to the feature image information of the one or more objects to be processed includes: in the event that the target processing authority allows processing a request for feature restoration processing for one or more objects to be processed, performing processing to restore the distinctive features for one or more objects to be processed in the video to be processed according to the image information of the features of one or more objects to be processed. 7. Video processing method applicable to the management platform, including: getting the original video; receiving a processing request for feature removal for the one or more objects to be processed, wherein the one or more objects to be processed is at least one object in the source video; according to a processing request for feature removal for one or more objects to be processed, obtaining feature image information corresponding to one or more objects to be processed, wherein the feature image information of one of the one or more objects to be processed contains position information pixels of all or part of the distinguishing features of one of one or more objects to be processed in the source video; And according to the feature image information of the one or more objects to be processed, performing feature removal processing on the one or more objects to be processed in the original video. 8. The method according to claim 7, wherein obtaining feature image information corresponding to one or more objects to be processed, in accordance with a processing request for removing features for one or more objects to be processed, includes: in accordance with a processing request to remove features for one or more objects to be processed, determining one or more identifiers of one or more objects to be processed, and obtaining feature image information corresponding to one or more identifiers of one or more objects to be processed. 9. The method of claim. 8, wherein receiving a processing request to remove features for one or more objects to be processed and determining one or more identifiers of one or more objects to be processed in accordance with the processing request to remove the features of one or more objects to be processed include: displaying images of objects present in the source video; receiving a processing request for feature removal for one or more objects to be processed, triggered by an operation of selecting at least one target image; obtaining at least one identifier of at least one target image as one or more identifiers of one or more objects to be processed; or receiving a processing request to remove features for one or more objects to be processed, wherein the processing request to remove distinctive features for one or more objects to be processed contains at least one image including one or more objects to be processed; for each of the one or more objects to be processed in the stored object images contained in the source video, determining a target image having a maximum similarity to the image containing the object to be processed, and obtaining an object identifier for the target image as identifiers of the objects to be processed; or displaying the original video; receiving a processing request for feature removal for one or more objects to be processed, triggered by an operation of selecting at least one area of the target image in the source video; according to the stored correspondence between object identifiers and image regions, determining at least one object identifier corresponding to at least one region of the target image as one or more identifiers of one or more objects to be processed. 10. The method of claim 8, wherein, before obtaining the feature image information corresponding to one or more identifiers of one or more objects to be processed, the method also includes: obtaining one or more identifiers of one or more objects to be processed; while obtaining image information of distinctive features corresponding to one or more identifiers of one or more objects to be processed includes: for each of the one or more objects to be processed, obtaining feature image information corresponding to both the object to be processed ID and the feature ID of the object to be processed. 11. The method according to claim 8 or 9, wherein obtaining feature image information corresponding to one or more identifiers of one or more objects to be processed includes: according to the stored correspondence between the object IDs in the source video and the object feature image information, determining the feature image information corresponding to the one or more IDs of the one or more objects to be processed; or transmitting one or more IDs of one or more objects to be processed to the image management terminal, so that the image management terminal determines feature image information corresponding to one or more IDs of one or more objects to be processed according to the stored correspondence between the object IDs in the original video and the image information distinctive features of objects; and receiving feature image information corresponding to one or more identifiers of one or more objects to be processed transmitted by the image control terminal. 12. Device for video processing, containing: a first video acquisition module configured to obtain a video to be processed, the video to be processed is obtained by performing feature removal processing on one or more objects in the original video; a first information acquiring module configured to receive a feature recovery processing request for one or more objects to be processed, wherein the one or more objects to be processed is one or more of the one or more objects for which the feature removal processing has been performed ; in accordance with a processing request for restoring features for one or more objects to be processed, obtaining feature image information corresponding to one or more objects to be processed, wherein the feature image information of one of the one or more objects to be processed contains position information pixels of all or part of the distinguishing features of one of one or more objects to be processed in the source video; And a feature recovery module configured to perform feature recovery processing for one or more objects to be processed in the video to be processed according to the feature image information of the one or more objects to be processed. 13. Device for video processing, containing: a second video acquisition module configured to acquire the original video; a second information acquisition module, configured to receive a processing request for feature removal for the one or more objects to be processed, wherein the one or more objects to be processed is at least one object in the original video; in accordance with a processing request to remove features for one or more objects to be processed, obtain feature image information corresponding to one or more objects to be processed, wherein the feature image information of one of the one or more objects to be processed contains position information pixels of all or part of the distinguishing features of one of one or more objects to be processed in the source video; And a feature removal module, configured to perform feature removal processing on the one or more objects to be processed in the source video according to the feature image information of the one or more objects to be processed. 14. A computer device containing a processor and a memory that stores at least one instruction, when loaded and executed by the processor, the video processing method according to any one of paragraphs is implemented. 1-11. 15. A computer-readable medium that stores at least one instruction, when loaded and executed by the processor, the video processing method according to any one of paragraphs is implemented. 1-11.

Images