Moving target identification method, device, storage medium and system
By acquiring the distance characteristics and global image characteristics of the moving target, determining the unoccluded area and sifting the local effective image, the problem of reduced recognition accuracy in multiple mobile scenes is solved, and higher recognition accuracy is achieved.
Patent Information
- Application Number
- CN202411177823.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-11-26
AI Technical Summary
In multiple mobile body scenes, mobile bodies may overlap or obstruct each other, resulting in reduced accuracy of real-time image recognition or unrecognized problems.
By acquiring the distance characteristics and global image characteristics of the moving target, determining the currently unoccluded area, filtering local valid images, and identifying them based on these images, fusing presets and historical images to supplement the complete outline of the target, and improving the recognition accuracy.
It effectively improves the recognition accuracy in multiple mobile scenes, and solves the problem of poor recognition accuracy due to overlap or obstruction.
Smart Images

Figure CN119027737A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of image recognition technology, and in particular to a moving target recognition method, device, storage medium and system. Background Art
[0002] In the prior art, there is a method for identifying a target object by performing pattern matching between a real-time captured image and a pre-stored image. In this method, in order to ensure the accuracy of pattern matching recognition, the real-time captured image should, as far as possible, be a global image that is free of external interference and contains a complete target object. However, in a scene with multiple moving objects, different moving objects may overlap or block each other during movement, which will greatly increase the interference factors in the real-time captured image. At this time, if one or more of the moving objects are used as the target recognition object, the recognition accuracy will be greatly reduced, or even the recognition will be unrecognizable. Summary of the invention
[0003] The main purpose of the present application is to provide a method, device, storage medium and system for identifying a moving target, so as to at least solve the problem in the prior art that in a scene with multiple moving objects, different moving objects may overlap or block each other during movement, and the recognition accuracy is poor when one or more of the moving objects are used as target recognition objects.
[0004] In order to achieve the above object, according to one aspect of the present application, a moving target recognition method is provided, the method comprising:
[0005] Acquire global image features and distance features of a moving target, wherein the global image features include the moving target, and the distance features of the moving target represent the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment;
[0006] In each acquisition cycle, the current unobstructed area of the moving target is determined according to the distance feature, wherein the distance feature and the global image feature of the moving target are acquired once in each acquisition cycle;
[0007] In each acquisition cycle, a local valid image corresponding to the current unblocked area is extracted from the global image features, and the moving target is identified based on the local valid image.
[0008] Optionally, acquiring the global image features and the distance features of the moving target includes: collecting the global image features and the distance features of the moving target at a preset period within a preset time period.
[0009] Optionally, the moving target includes a first part and a second part, and in each acquisition cycle, the current unobstructed area of the moving target is determined according to the distance feature, including: when the first part of the moving target in the global image feature is obstructed by a preset object and the second part of the moving target in the global image feature is not obstructed by the preset object, determining the second part as the current unobstructed area of the moving target; when the first part of the moving target in the global image feature is not obstructed by the preset object and the second part of the moving target in the global image feature is obstructed by the preset object, determining the first part as the current unobstructed area of the moving target.
[0010] Optionally, the reference image includes a preset complete image and a historical local image, the preset complete image is a repaired image relative to a previous frame of the repaired image, and the historical local image is a local valid image relative to a previous frame of the repaired image, and identifying the moving target based on the local valid image includes: fusing all local valid images within a preset time period to obtain a repaired image, and extracting a repair contour based on the repaired image; determining the complete contour of the moving target according to all distance features within the preset time period; determining the coverage value of the repair contour according to the complete contour and the repair contour; and executing a first preset method or a second preset method according to the range of the coverage value of the repair contour to identify the moving target, the first preset method being a comparative identification between the repaired image and the preset complete image, and the second preset method being a comparative identification between the repaired image and the historical local image.
[0011] Optionally, fusing all the local valid images within a preset time period to obtain the repaired image includes: splicing all the local valid images within the preset time period to obtain the repaired image.
[0012] Optionally, determining the coverage value of the repair contour according to the complete contour and the repair contour includes: determining the coverage value of the repair contour according to P=w / W, wherein P is the coverage value of the repair contour, w is the area of the repair contour, and W is the area of the complete contour.
[0013] Optionally, according to the range of the coverage value of the repair contour, the first preset method is executed, or the second preset method is executed, including: when the coverage value of the repair contour is greater than or equal to a preset threshold, the first preset method is executed; when the coverage value of the repair contour is less than the preset threshold, the second preset method is executed.
[0014] According to another aspect of the present application, a moving target recognition device is provided, the device comprising:
[0015] An acquisition unit, used to acquire global image features and distance features of a moving target, wherein the global image features include the moving target, and the distance features of the moving target represent the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment;
[0016] A determination unit, configured to determine, in each acquisition cycle, a current unobstructed area of the moving target according to the distance feature, wherein the distance feature and the global image feature of the moving target are acquired once in each acquisition cycle;
[0017] A processing unit is used to extract a local valid image corresponding to the current unobstructed area from the global image features in each acquisition cycle, and identify the moving target based on the local valid image.
[0018] According to another aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute any one of the moving target recognition methods.
[0019] According to another aspect of the present application, a moving target recognition system is provided, the system comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include means for executing any one of the moving target recognition methods.
[0020] By applying the technical solution of the present application, the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment, as well as the global image features, is obtained, and in each acquisition cycle, the current unobstructed area of the moving target is determined according to the distance features, thereby facilitating the screening of local valid images, and thus identifying the moving target based on the local valid images, thereby improving the accuracy of recognition, and further solving the problem in the prior art that in a scene with multiple moving objects, different moving objects may overlap or obstruct each other during the movement, and the recognition accuracy is poor when one or more of the moving objects are used as target recognition objects. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] The drawings constituting part of the present application are used to provide a further understanding of the present application. The exemplary embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0022] Figure 1A schematic diagram of a flow chart of a moving target recognition method provided according to an embodiment of the present application is shown;
[0023] Figure 2 A schematic diagram showing a current unblocked area provided in an embodiment of the present application is shown;
[0024] Figure 3 A schematic diagram of a flow chart of another moving target recognition method provided according to an embodiment of the present application is shown;
[0025] Figure 4 A structural block diagram of a moving target recognition device provided according to an embodiment of the present application is shown. DETAILED DESCRIPTION
[0026] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.
[0027] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.
[0028] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0029] As introduced in the background technology, there is a method in the prior art to identify the target object by performing pattern matching between the real-time acquired image and the pre-stored image. In this method, in order to ensure the accuracy of pattern matching recognition, the real-time acquired image should, as far as possible, be a global image that is free of external interference and contains the complete target object. However, in a scene with multiple moving bodies, different moving bodies may overlap or block each other during the movement, which will greatly increase the interference factors in the real-time acquired image. At this time, if one or more of the moving bodies are used as target recognition objects, the recognition accuracy will be greatly reduced or even unrecognizable. In order to solve the problem that in the prior art, different moving bodies may overlap or block each other during the movement in a scene with multiple moving bodies, and the recognition accuracy is poor when one or more of the moving bodies are used as target recognition objects, the embodiments of the present application provide a moving target recognition method, device, storage medium and system.
[0030] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.
[0031] In this embodiment, a method for identifying a moving target running on a mobile terminal, a computer terminal or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0032] Figure 1 FIG. 1 is a flow chart of a moving target recognition method provided according to an embodiment of the present application. Figure 1 As shown, the method comprises the following steps:
[0033] Step S101, obtaining global image features and distance features of a moving target, wherein the global image features include the moving target, and the distance features of the moving target represent the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment;
[0034] Step S101, i.e. acquiring global image features and distance features of moving targets, includes: acquiring the global image features and the distance features of moving targets at a preset period within a preset time period.
[0035] Specifically, the global image features and the distance features of the moving target may be collected within 10 seconds with 100 ms as a preset period.
[0036] Step S102, in each acquisition cycle, determine the current unobstructed area of the moving target according to the distance feature, and the pixel distance of the significant feature of the known position from the target in the image, and then correct the two objects to the actual scene through perspective transformation to obtain the corresponding motion feature. The distance feature and global image feature of the moving target are obtained once in each acquisition cycle;
[0037] In one embodiment of the present application, the moving target includes a first part and a second part. In each acquisition cycle, the current unobstructed area of the moving target is determined according to the distance feature, including: when the first part of the moving target in the global image feature is obstructed by a preset object, and the second part of the moving target in the global image feature is not obstructed by the preset object, determining the second part as the current unobstructed area of the moving target; when the first part of the moving target in the global image feature is not obstructed by the preset object, and the second part of the moving target in the global image feature is obstructed by the preset object, determining the first part as the current unobstructed area of the moving target.
[0038] Specifically, refer to Figure 2 As shown in the left figure, during the T1 acquisition period, the moving target a is partially blocked by object b, and the L1 area of the moving target a is the current unblocked area; refer to Figure 2 As shown in the right figure, during the T2 acquisition period, a part of the moving target a is blocked by the object b, and the L2 area of the moving target a is the current unblocked area.
[0039] Step S103, in each acquisition cycle, extract the local effective image corresponding to the current unobstructed area from the global image features. By calculating the effective scale of the obstructed edge area (i.e., the largest range), the effective image is extracted through image processing technology. And the moving target is identified based on the local effective image.
[0040] In the above steps, the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment, as well as the global image features, are obtained, and in each acquisition cycle, the current unobstructed area of the moving target is determined according to the distance features, so as to facilitate the screening of local valid images, thereby identifying the moving target based on the local valid images, thereby improving the accuracy of recognition, and further solving the problem in the prior art that in a scene with multiple moving bodies, different moving bodies may overlap or obstruct each other during the movement, and the recognition accuracy is poor when one or more of the moving bodies are used as target recognition objects.
[0041] The image acquisition of the target to be identified is integrated with the distance acquisition. Specifically, the occluded and unoccluded parts of the target to be identified are determined through distance acquisition. Then, a local valid image is extracted from the image acquisition according to the unoccluded part. Multiple local valid images are integrated to obtain a repaired image. This can maximize the overall features of the target to be identified, thereby effectively improving the recognition accuracy. In addition, the local valid images carrying the classification mark are stored after classification, so that the present application can also use some local features to complete target recognition, thereby realizing the recognition of moving targets by combining distance and image.
[0042] The reference image includes a preset complete image and a historical partial image, wherein the preset complete image is a restored image of a previous frame relative to the restored image, and the historical partial image is a local valid image of a previous frame relative to the restored image;
[0043] Because the image at each moment can be fused with the target image in the previous and next time periods, the complete image of the target at any moment can be supplemented. Therefore, the motion profile of the target in any specified time period can also be completely fused.
[0044] The step S103 of identifying the moving target based on the local effective image includes:
[0045] Fusing all local valid images within a preset time period to obtain a repaired image, and extracting a repair contour based on the repaired image;
[0046] Specifically, all the above-mentioned local valid images within the above-mentioned preset time period are spliced to obtain the above-mentioned repaired image. The starting coordinate position of each image is corrected according to the coordinates of the pixels and the geometric size of the target object. Because the camera is fixed in the production scene, the z and y coordinate axes of the target object are relatively fixed. Therefore, this step often only needs to correct the coordinates in one direction, such as the x-axis.
[0047] Determine the complete outline of the moving target based on all distance features within the preset time period;
[0048] Determine the coverage value of the repaired contour according to the complete contour and the repaired contour;
[0049] Specifically, the coverage value of the repair contour is determined based on the complete contour and the repair contour, including: determining the coverage value of the repair contour based on P=w / W, wherein P is the coverage value of the repair contour, w is the area of the repair contour, and W is the area of the complete contour.
[0050] According to the range of coverage values of the above-mentioned repair contour, the first preset method or the second preset method is executed to identify the above-mentioned moving target. The above-mentioned first preset method is to compare and identify the above-mentioned repaired image with the above-mentioned preset complete image, and the above-mentioned second preset method is to compare and identify the above-mentioned repaired image with the above-mentioned historical local image.
[0051] Specifically, when the coverage value of the above-mentioned repair contour is greater than or equal to the preset threshold, the above-mentioned first preset method is executed; when the coverage value of the above-mentioned repair contour is less than the above-mentioned preset threshold, the above-mentioned second preset method is executed.
[0052] By calculating the coverage value, it is selected whether to compare and identify with a preset complete image or with a historical partial image. The accuracy of comparing and identifying with a historical partial image is higher.
[0053] The preset threshold is preferably 0.6, which means that when the restored image reaches more than 60% of the complete image of the moving target, the comparison and identification between the restored image and the preset complete image is performed.
[0054] What is being compared is the color difference and brightness difference between pixels. Due to the existence of light and shadow and other factors, the same object or the same part may show large differences in different coordinates. Corresponding comparison and correction are needed so that the image will not appear abrupt.
[0055] Taking the RGBA data format as an example, the R, G, B, and A values of each pixel in the repaired image are compared with the corresponding RGBA values of the pixel in the preset complete image, and the difference ratio in each dimension is calculated. If the ratio is greater than a set threshold, the color value of the pixel is corrected according to the brightness value of the surrounding image.
[0056] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, the implementation process of the moving target recognition method of the present application will be described in detail below in conjunction with specific embodiments.
[0057] This embodiment relates to a specific moving target recognition method, such as Figure 3 As shown, the following steps are included:
[0058] Step S1: collecting global image features and distance features of moving targets at a preset period within a preset time period;
[0059] Step S2: the moving target includes a first part and a second part. In each acquisition cycle, the current unobstructed area of the moving target is determined according to the distance feature. When the first part of the moving target in the global image feature is obstructed by a preset object and the second part of the moving target in the global image feature is not obstructed by the preset object, the second part is determined to be the current unobstructed area of the moving target; when the first part of the moving target in the global image feature is not obstructed by the preset object and the second part of the moving target in the global image feature is obstructed by the preset object, the first part is determined to be the current unobstructed area of the moving target;
[0060] Step S3: In each acquisition cycle, extract the local valid image corresponding to the current unobstructed area from the global image features.
[0061] Step S4: fusing all local valid images within a preset time period to obtain a repaired image, and extracting a repaired contour based on the repaired image; determining a complete contour of the moving target according to all distance features within the preset time period; and determining a coverage value of the repaired contour according to the complete contour and the repaired contour;
[0062] Step S5: According to the range of the coverage value of the repaired contour, execute the first preset method or the second preset method to identify the moving target. The first preset method is to compare and identify the repaired image with the preset complete image, and the second preset method is to compare and identify the repaired image with the historical local image.
[0063] By acquiring the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment, as well as the global image features, and in each acquisition cycle, determining the current unobstructed area of the moving target according to the distance features, it is convenient to screen out local valid images, thereby identifying the moving target based on the local valid images, improving the accuracy of recognition, and further solving the problem in the prior art that in a scene with multiple moving objects, different moving objects may overlap or obstruct each other during the movement, and the recognition accuracy is poor when one or more of the moving objects are used as target recognition objects.
[0064] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0065] The present application embodiment also provides a moving target identification device. It should be noted that the moving target identification device of the present application embodiment can be used to execute the moving target identification method provided by the present application embodiment. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions thereof will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware is also possible and contemplated.
[0066] The following is an introduction to the moving target recognition device provided in the embodiments of the present application.
[0067] Figure 4 is a structural block diagram of a moving target recognition device provided according to an embodiment of the present application. Figure 4 As shown, the device comprises:
[0068] An acquisition unit 41 is used to acquire global image features and distance features of a moving target, wherein the global image features include the moving target, and the distance features of the moving target represent the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment;
[0069] A determination unit 42 is used to determine the current unobstructed area of the moving target according to the distance feature in each acquisition cycle, wherein the distance feature and the global image feature of the moving target are acquired once in each acquisition cycle;
[0070] The processing unit 43 is used to extract the local effective image corresponding to the current unblocked area from the global image features in each acquisition cycle, and identify the moving target based on the local effective image.
[0071] In the above-mentioned device, by obtaining the distance moved by the above-mentioned moving target from the moment when the current acquisition cycle starts to the current moment, as well as the global image features, and in each acquisition cycle, determining the current unobstructed area of the above-mentioned moving target according to the above-mentioned distance features, it is convenient to screen out local valid images, and thus identify the moving target based on the local valid images, thereby improving the accuracy of recognition, and further solving the problem in the prior art that in a scene with multiple moving bodies, different moving bodies may overlap or obstruct each other during the movement process, and the recognition accuracy is poor when one or more of the moving bodies are used as target recognition objects.
[0072] In one embodiment of the present application, the acquisition unit includes an acquisition module, and the acquisition module is used to collect the above-mentioned global image features and the above-mentioned distance features of the above-mentioned moving target at a preset period within a preset time period.
[0073] In one embodiment of the present application, the moving target includes a first part and a second part. In each acquisition cycle, the determination unit includes a first determination module and a second determination module. The first determination module is used to determine that the second part is the current unobstructed area of the moving target when the first part of the moving target in the global image feature is blocked by a preset object and the second part of the moving target in the global image feature is not blocked by the preset object; the second determination module is used to determine that the first part is the current unobstructed area of the moving target when the first part of the moving target in the global image feature is not blocked by the preset object and the second part of the moving target in the global image feature is blocked by the preset object.
[0074] In one embodiment of the present application, the reference image includes a preset complete image and a historical local image, the preset complete image is a repaired image relative to the previous frame of the repaired image, and the historical local image is a local valid image relative to the previous frame of the repaired image, and the processing unit includes a first processing module, a third determination module, a fourth determination module and a second processing module, the first processing module is used to fuse all local valid images within a preset time period to obtain a repaired image, and extract a repair contour based on the repaired image; the third determination module is used to determine the complete contour of the moving target according to all distance features within the preset time period; the fourth determination module is used to determine the coverage value of the repair contour according to the complete contour and the repair contour; the second processing module is used to execute the first preset method or the second preset method according to the range of the coverage value of the repair contour to identify the moving target, the first preset method is the comparison and identification of the repaired image and the preset complete image, and the second preset method is the comparison and identification of the repaired image and the historical local image.
[0075] In one embodiment of the present application, the first processing module includes a first processing submodule, and the first processing submodule is used to splice all the above-mentioned local valid images within the above-mentioned preset time period to obtain the above-mentioned repaired image.
[0076] In one embodiment of the present application, the fourth determination module includes a determination submodule, which is used to determine the coverage value of the above-mentioned repair contour according to P=w / W, wherein P is the coverage value of the above-mentioned repair contour, w is the area of the above-mentioned repair contour, and W is the area of the above-mentioned complete contour.
[0077] In one embodiment of the present application, the second processing module includes a second processing sub-module and a third processing sub-module. The second processing sub-module is used to execute the above-mentioned first preset method when the coverage value of the above-mentioned repair contour is greater than or equal to the preset threshold; the third processing sub-module is used to execute the above-mentioned second preset method when the coverage value of the above-mentioned repair contour is less than the above-mentioned preset threshold.
[0078] The moving target recognition device includes a processor and a memory. The acquisition unit, the determination unit, the processing unit, etc. are all stored in the memory as program units, and the processor executes the program units stored in the memory to implement corresponding functions. The modules are all located in the same processor; or, the modules are located in different processors in any combination.
[0079] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and the kernel parameters are adjusted to solve the problem that in a scene with multiple moving objects, different moving objects may overlap or block each other during movement, and the recognition accuracy is poor when one or more moving objects are used as target recognition objects.
[0080] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.
[0081] An embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the moving target recognition method.
[0082] An embodiment of the present invention provides a processor, and the processor is used to run a program, wherein the moving target recognition method is executed when the program is run.
[0083] An embodiment of the present invention provides a device, the device includes a processor, a memory, and a program stored in the memory and executable on the processor, and when the processor executes the program, at least the following steps are implemented: obtaining global image features and distance features of a moving target, wherein the moving target is included in the global image features, and the distance features of the moving target represent the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment; in each acquisition cycle, determining the current unobstructed area of the moving target based on the distance features, wherein the distance features and global image features of the moving target are obtained once in each acquisition cycle; in each acquisition cycle, extracting a local valid image corresponding to the current unobstructed area from the global image features, and identifying the moving target based on the local valid image. The device in this article may be a server, a PC, a PAD, a mobile phone, etc.
[0084] Optionally, acquiring the global image features and the distance features of the moving target includes: collecting the global image features and the distance features of the moving target at a preset period within a preset time period.
[0085] Optionally, the moving target includes a first part and a second part, and in each acquisition cycle, the current unobstructed area of the moving target is determined according to the distance feature, including: when the first part of the moving target in the global image feature is obstructed by a preset object, and the second part of the moving target in the global image feature is not obstructed by the preset object, determining the second part as the current unobstructed area of the moving target; when the first part of the moving target in the global image feature is not obstructed by the preset object, and the second part of the moving target in the global image feature is obstructed by the preset object, determining the first part as the current unobstructed area of the moving target.
[0086] Optionally, the reference image includes a preset complete image and a historical local image, the preset complete image is a repaired image relative to the previous frame of the repaired image, and the historical local image is a local valid image relative to the previous frame of the repaired image, and identifying the moving target based on the local valid image includes: fusing all local valid images within a preset time period to obtain a repaired image, and extracting a repair contour based on the repaired image; determining the complete contour of the moving target according to all distance features within the preset time period; determining the coverage value of the repair contour according to the complete contour and the repair contour; according to the range of the coverage value of the repair contour, executing the first preset method or the second preset method to identify the moving target, the first preset method is a comparative identification between the repaired image and the preset complete image, and the second preset method is a comparative identification between the repaired image and the historical local image.
[0087] Optionally, fusing all local valid images within a preset time period to obtain the repaired image includes: splicing all the above local valid images within the preset time period to obtain the above repaired image.
[0088] Optionally, determining the coverage value of the repair contour based on the complete contour and the repair contour includes: determining the coverage value of the repair contour based on P=w / W, wherein P is the coverage value of the repair contour, w is the area of the repair contour, and W is the area of the complete contour.
[0089] Optionally, according to the range of the coverage value of the above-mentioned repair contour, the first preset method is executed, or the second preset method is executed, including: when the coverage value of the above-mentioned repair contour is greater than or equal to the preset threshold, the first preset method is executed; when the coverage value of the above-mentioned repair contour is less than the above-mentioned preset threshold, the second preset method is executed.
[0090] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that is initialized with at least the following method steps: obtaining global image features and distance features of a moving target, the above-mentioned global image features include the above-mentioned moving target, and the distance features of the above-mentioned moving target represent the distance moved by the above-mentioned moving target from the moment when the current acquisition cycle starts to the current moment; in each acquisition cycle, determining the current unobstructed area of the above-mentioned moving target based on the above-mentioned distance features, wherein the distance features and global image features of the moving target are obtained once in each of the above-mentioned acquisition cycles; in each acquisition cycle, extracting a local valid image corresponding to the above-mentioned current unobstructed area from the above-mentioned global image features, and identifying the above-mentioned moving target based on the above-mentioned local valid image.
[0091] Optionally, acquiring the global image features and the distance features of the moving target includes: collecting the global image features and the distance features of the moving target at a preset period within a preset time period.
[0092] Optionally, the moving target includes a first part and a second part, and in each acquisition cycle, the current unobstructed area of the moving target is determined according to the distance feature, including: when the first part of the moving target in the global image feature is obstructed by a preset object, and the second part of the moving target in the global image feature is not obstructed by the preset object, determining the second part as the current unobstructed area of the moving target; when the first part of the moving target in the global image feature is not obstructed by the preset object, and the second part of the moving target in the global image feature is obstructed by the preset object, determining the first part as the current unobstructed area of the moving target.
[0093] Optionally, the reference image includes a preset complete image and a historical local image, the preset complete image is a repaired image relative to the previous frame of the repaired image, and the historical local image is a local valid image relative to the previous frame of the repaired image, and identifying the moving target based on the local valid image includes: fusing all local valid images within a preset time period to obtain a repaired image, and extracting a repair contour based on the repaired image; determining the complete contour of the moving target according to all distance features within the preset time period; determining the coverage value of the repair contour according to the complete contour and the repair contour; according to the range of the coverage value of the repair contour, executing the first preset method or the second preset method to identify the moving target, the first preset method is a comparative identification between the repaired image and the preset complete image, and the second preset method is a comparative identification between the repaired image and the historical local image.
[0094] Optionally, fusing all local valid images within a preset time period to obtain the repaired image includes: splicing all the above local valid images within the preset time period to obtain the above repaired image.
[0095] Optionally, determining the coverage value of the repair contour based on the complete contour and the repair contour includes: determining the coverage value of the repair contour based on P=w / W, wherein P is the coverage value of the repair contour, w is the area of the repair contour, and W is the area of the complete contour.
[0096] Optionally, according to the range of the coverage value of the above-mentioned repair contour, the first preset method is executed, or the second preset method is executed, including: when the coverage value of the above-mentioned repair contour is greater than or equal to the preset threshold, the first preset method is executed; when the coverage value of the above-mentioned repair contour is less than the above-mentioned preset threshold, the second preset method is executed.
[0097] The present application also provides a moving target recognition system, which includes: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing any of the above-mentioned moving target recognition methods. By obtaining the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment, and the global image features, and in each acquisition cycle, the current unobstructed area of the moving target is determined according to the distance features, so as to facilitate the screening of local valid images, thereby identifying the moving target based on the local valid images, improving the accuracy of recognition, and thus solving the problem that in the prior art, in a scene with multiple moving bodies, different moving bodies may overlap or obstruct each other during movement, and the recognition accuracy is poor when one or more of the moving bodies are used as target recognition objects.
[0098] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.
[0099] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0100] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0101] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0102] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0103] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0104] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0105] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0106] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0107] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:
[0108] 1) The moving target recognition method of the present application obtains the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment, as well as the global image features, and in each acquisition cycle, determines the current unobstructed area of the moving target according to the distance features, thereby facilitating the screening of local valid images, thereby identifying the moving target based on the local valid images, improving the recognition accuracy, and further solving the problem in the prior art that in a scene with multiple moving objects, different moving objects may overlap or obstruct each other during the movement process, and the recognition accuracy is poor when one or more of the moving objects are used as target recognition objects.
[0109] 2) The moving target recognition device of the present application obtains the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment, as well as the global image features, and in each acquisition cycle, determines the current unobstructed area of the moving target according to the distance features, thereby facilitating the screening of local valid images, and thus identifying the moving target based on the local valid images, thereby improving the recognition accuracy, and further solving the problem in the prior art that in a scene with multiple moving bodies, different moving bodies may overlap or obstruct each other during the movement, and the recognition accuracy is poor when one or more of the moving bodies are used as target recognition objects.
[0110] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims
1. A moving target recognition method, characterized in that: include: Acquire global image features and distance features of a moving target, wherein the global image features include the moving target, and the distance features of the moving target represent the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment; In each acquisition cycle, the current unobstructed area of the moving target is determined according to the distance feature, wherein the distance feature and the global image feature of the moving target are acquired once in each acquisition cycle; In each acquisition cycle, a local valid image corresponding to the current unblocked area is extracted from the global image features, and the moving target is identified based on the local valid image.
2. The method according to claim 1, characterized in that: Obtain global image features and distance features of moving targets, including: The global image features and the distance features of the moving target are collected at a preset period within a preset time period.
3. The method according to claim 1, characterized in that The moving target includes a first part and a second part. In each acquisition cycle, determining a current unblocked area of the moving target according to the distance feature includes: When a first portion of the moving target in the global image feature is blocked by a preset object and a second portion of the moving target in the global image feature is not blocked by the preset object, determining the second portion as a current unblocked area of the moving target; When a first portion of the moving target in the global image feature is not blocked by a preset object and a second portion of the moving target in the global image feature is blocked by the preset object, the first portion is determined to be a current unblocked area of the moving target.
4. The method according to claim 1, characterized in that: The reference image includes a preset complete image and a historical partial image, wherein the preset complete image is a repaired image relative to a previous frame of the repaired image, and the historical partial image is a local valid image relative to a previous frame of the repaired image. Identifying the moving target based on the local valid image includes: Fusing all local valid images within a preset time period to obtain a repaired image, and extracting a repair contour based on the repaired image; Determine the complete outline of the moving target according to all distance features within the preset time period; Determining a coverage value of the repaired contour according to the complete contour and the repaired contour; According to the range of the coverage value of the repaired contour, the first preset method or the second preset method is executed to identify the moving target. The first preset method is to compare and identify the repaired image with the preset complete image, and the second preset method is to compare and identify the repaired image with the historical local image.
5. The method according to claim 4, characterized in that All local valid images within a preset time period are integrated to obtain a repaired image, including: All the local valid images within the preset time period are spliced to obtain the repaired image.
6. The method according to claim 4, characterized in that Determining a coverage value of the repaired contour according to the complete contour and the repaired contour includes: According to P=w / W, the coverage value of the repaired contour is determined, wherein P is the coverage value of the repaired contour, w is the area of the repaired contour, and W is the area of the complete contour.
7. The method according to claim 4, characterized in that According to the range of the coverage value of the repair contour, the first preset method is executed, or the second preset method is executed, including: When the coverage value of the repaired contour is greater than or equal to a preset threshold, executing the first preset method; When the coverage value of the repaired contour is less than the preset threshold, the second preset mode is executed.
8. A moving target recognition device, characterized in that: include: An acquisition unit, used to acquire global image features and distance features of a moving target, wherein the global image features include the moving target, and the distance features of the moving target represent the distance moved by the moving target from the moment when the current acquisition cycle starts to the current moment; A determination unit, configured to determine, in each acquisition cycle, a current unobstructed area of the moving target according to the distance feature, wherein the distance feature and the global image feature of the moving target are acquired once in each acquisition cycle; A processing unit is used to extract a local valid image corresponding to the current unobstructed area from the global image features in each acquisition cycle, and identify the moving target based on the local valid image.
9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the moving target recognition method according to any one of claims 1 to 7.
10. A moving target recognition system, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing the motion target recognition method described in any one of claims 1 to 7.
