Pedestrian red light running recognition and snapshot method and system, electronic equipment and storage medium
Through the method of detecting targets by video stream data acquisition and deep learning model, combined with path tracking and direction judgment, a pedestrian running a red light recognition and capture method was established, which solved the problem of low recognition accuracy in the existing technology, and achieved higher recognition accuracy and accurate capture effect.
Patent Information
- Application Number
- CN202510028450.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-08
- Publication Date
- 2025-04-25
AI Technical Summary
The prior art is not very accurate when identifying pedestrians running red lights, and is prone to misjudgment, especially when pedestrian traffic lights are changed or when there are special staff.
Through steps such as video stream data acquisition, target detection, target filtering and extraction, trajectory tracking and behavioral judgment, a pedestrian identification and capture method is established. This method uses a deep learning model to detect targets, combines path tracking and direction judgment to reduce misjudgment.
It improves the accuracy of pedestrians' identification of running red lights, reduces misjudgments, and accurately captures pedestrians running red lights, avoiding misjudgments and misjudgments.
Smart Images
Figure CN119888644A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of recognition and capture of pedestrians running red lights, and in particular to a method, system, electronic equipment and storage medium for recognition and capture of pedestrians running red lights. Background Art
[0002] In factories or industrial parks, at crossroads or intersections where pedestrians and vehicles mix, traffic lights will be installed to regulate the behavior of pedestrians and vehicles in order to ensure safety and traffic efficiency. For the identification and management of pedestrians and vehicles that run red lights, technology + manual methods are generally used. On the one hand, the system demarcates the driving and pedestrian identification areas. According to the current status of the traffic light, if it is a green light for pedestrians, it determines whether the vehicle appears in the driving identification area. If so, it is determined that the vehicle has run a red light; if it is a red light for pedestrians, it determines whether there are pedestrians in the area. If so, it is determined that the pedestrian has run a red light. Furthermore, by identifying the license plate of the vehicle, the vehicle that runs the red light can be punished accordingly; by identifying the face of the pedestrian who runs the red light or manual judgment, the specific person who runs the red light can be found and the corresponding punishment education can be given to him.
[0003] Most of the red light running captures on the market are currently aimed at vehicles. The general method is to determine the target through sensors and then capture it through cameras. Specifically, when the front wheel of the vehicle passes over the induction coil on the ground, the camera takes the first photo; when the rear wheel of the vehicle passes over the induction coil, the second photo is taken; when the rear wheel of the entire vehicle passes through the intersection and passes over the induction coil on the opposite side, the third photo is taken. These three photos respectively record the position movement of the vehicle during the red light process, thereby confirming whether the vehicle has run a red light. However, this method is quite different from identifying pedestrians running red lights, and cannot be directly applied to the identification of pedestrians running red lights.
[0004] Among the methods for identifying pedestrians running red lights, most of them adopt direct identification of pedestrians and then take snapshots based on the status of the traffic lights. That is, the system analyzes the movable objects of the camera in real time. If a person is found and the current status of the pedestrian light is red, the snapshot is taken. However, this algorithm has certain defects and the accuracy of pedestrian red light identification is not high, which is specifically manifested in the following aspects: a. When the pedestrian traffic light is green, but in the last few seconds, the pedestrian enters the zebra crossing in a normal way, but has not yet completely passed, the traffic light turns red. In this case, the system will mistakenly judge the pedestrian as a red light runner; b. In addition to pedestrians passing through traffic lights, there are some special workers at the intersection, such as cleaners. These people did not run a red light, but the system may also judge them as running a red light; c. In some special places, there are people who operate special work objects, such as forklift drivers and open-top truck drivers. They operate according to driving rules, but the system will judge that the driver is a pedestrian and runs a red light.
[0005] Some solutions also set a transition time (e.g. 5 seconds) for the red light signal switch. When the light switches to red, the system will check whether there are pedestrians still on the zebra crossing during this transition time. If so, it will not be judged as running a red light for the time being. However, although the transition time (5 seconds) is introduced, there is still a possibility of misjudgment: if the passage area is long and the pedestrian speed is slow, the pedestrian cannot pass through the passage area within 5 seconds, and the pedestrian will still be judged as running a red light; if the passage area is short or the pedestrian speed is fast, after the light switches to red, the pedestrian quickly passes through the passage area within 5 seconds, and the system will think that the pedestrian is passing normally and will not judge that he ran a red light; and for the above two situations b and c, misjudgment will still occur. Summary of the invention
[0006] The purpose of the present invention is to provide a method, system, electronic device and storage medium for identifying and capturing pedestrians running red lights, which can reduce misjudgments and improve the accuracy of identifying pedestrians running red lights.
[0007] According to a first aspect of the present invention, a method for identifying and capturing pedestrians running red lights is provided, comprising: Video stream data acquisition: collect video data and split the video into video frame images; Target detection: Detect each frame of the image to obtain the bounding boxes, bounding box coordinates and time of human targets and non-human targets; the human targets include pedestrian targets and non-pedestrian targets; Target filtering and extraction: Analyze and filter out non-human targets and non-pedestrian targets to extract pedestrian targets; Trajectory tracking: establish trajectory objects, determine the relationship between characters in each frame of the image, and form a tracking trajectory; Behavior judgment: Calculate the direction of the pedestrian target based on the tracking trajectory; obtain the time when the pedestrian target enters the zebra crossing and the time when the traffic light switches to red; determine whether the pedestrian runs a red light; Capture pictures: capture images of pedestrians running red lights and save them; Output result information: Get the time when the image of the pedestrian running a red light is taken, and output it together with the image.
[0008] According to the method for identifying and capturing pedestrians running red lights, in the target filtering and extraction step, non-human targets and non-pedestrian targets are analyzed and filtered out to extract pedestrian targets, including: According to the frame coordinates, the frames of human targets that are contained in the non-human frame or half-enclosed in the non-human frame are determined, and the frames of these human targets are filtered out; Filter out all non-pedestrian bounding boxes; The objects corresponding to the remaining bounding boxes are extracted as pedestrian objects.
[0009] According to the method for identifying and capturing pedestrians running red lights, in the trajectory tracking step, a trajectory object is established, and the character relationship of each frame image is determined to form a tracking trajectory, which includes: Establish trajectory points for pedestrian targets in the image; According to the time sequence, the pedestrian target in the current video frame image is cyclically compared with the pedestrian target in the previous video frame image to determine whether the pedestrian target appears for the first time; if it is not the first time, the trajectory points of the pedestrian target in the two video frames are associated to form the pedestrian's trajectory line.
[0010] According to the method for identifying and capturing pedestrians running red lights, in the trajectory tracking step, the content of judging whether the pedestrian target appears for the first time includes: Cyclic comparison of the overlapping degree of the pedestrian target in the current video frame image and the border of the pedestrian target in the previous video frame image; predicting the position of the pedestrian target based on the Kalman filter prediction method; A first threshold is set for the degree of overlap. If the degree of overlap is greater than or equal to the first threshold, and the difference between the predicted value and the actual value of the pedestrian target position is less than or equal to the second threshold, it is judged as the same pedestrian target that does not appear for the first time; otherwise, it is judged as the first appearance.
[0011] According to the method for identifying and capturing pedestrians running red lights, in the behavior judgment step, Calculating the direction of travel of the pedestrian target according to the tracking trajectory includes: setting a third threshold for the number of trajectory points, and when the number of trajectory points in the trajectory line is greater than or equal to the third threshold, taking the extension direction of the trajectory line as the direction of travel of the pedestrian target; Acquiring the time when the pedestrian target enters the zebra crossing includes: acquiring the time of the video frame image captured when the pedestrian enters the zebra crossing, and the time is used as the time when the pedestrian target enters the zebra crossing.
[0012] According to the method for identifying and capturing pedestrians running red lights, 0.75≤first threshold≤0.85, 10 pixels≤second threshold≤20 pixels, and 3≤third threshold.
[0013] According to the method for identifying and capturing pedestrians running red lights, in the behavior judgment step, judging whether a pedestrian has run a red light includes: Determine whether the pedestrian target is moving in the direction of crossing the zebra crossing; if not, remove the pedestrian target; Determine whether the traffic light is green when the pedestrian enters the zebra crossing; if so, remove the pedestrian; The remaining pedestrians that were not eliminated were targets of pedestrians running red lights.
[0014] According to a second aspect of the present invention, a pedestrian red light violation recognition and capture system is provided, comprising a processor and a memory, wherein the memory stores a plurality of instructions; the processor loads instructions from the memory to execute the pedestrian red light violation recognition and capture method as described above.
[0015] According to a third aspect of the present invention, an electronic device is provided, comprising a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method for identifying and capturing pedestrians running red lights as described above.
[0016] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method for identifying and capturing pedestrians running red lights as described above are implemented.
[0017] Beneficial effects: The deep learning model is used to detect the target, and the path tracking, direction determination and other methods are used to determine the target. In combination with real-time video, it can accurately capture people who run red lights. Taking into account the influence of other pedestrians in public places, the moving direction of pedestrians is detected by tracking their trajectories, excluding those who enter the zebra crossing but do not intend to pass the traffic light, filtering out the interference of other behaviors on the capture, so as to reduce misjudgment and improve the accuracy of pedestrian red light running recognition; By including people and objects, the interference of people on non-human targets on recognition is eliminated, and the detection and positioning of pedestrians are more accurate, so as to reduce misjudgment and improve the accuracy of pedestrian red light running recognition; The defects of the time transition method are overcome. By using the trajectory tracking method, as long as the time when a person enters the zebra crossing and the status of the traffic light at that time are determined, it is possible to accurately determine whether the person is a red light runner, thereby avoiding misjudgment and missed judgment and improving the accuracy of pedestrian red light running recognition.
[0018] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] The present invention is further described below in conjunction with the accompanying drawings and embodiments; Figure 1 A simplified flow chart of an embodiment of the present invention; Figure 2 A network diagram of an embodiment of the present invention. DETAILED DESCRIPTION
[0020] This section will describe in detail the specific embodiments of the present invention. The preferred embodiments of the present invention are shown in the accompanying drawings. The purpose of the accompanying drawings is to supplement the description of the text part of the specification with graphics, so that people can intuitively and vividly understand each technical feature and the overall technical solution of the present invention, but it cannot be understood as a limitation on the scope of protection of the present invention.
[0021] Reference Figure 1-2 The method for identifying and capturing a pedestrian running a red light according to an embodiment of the present invention comprises the following steps: Video stream data acquisition: collect video data and split the video into video frame images; Target detection: Load the deep learning model to perform target detection, detect each frame of the image, and obtain the bounding box, bounding box coordinates, and time (i.e., the time when the frame was taken) of human targets and non-human targets; the human targets include pedestrian targets and non-pedestrian targets; Target filtering and extraction: Analyze and filter out non-human targets and non-pedestrian targets to extract pedestrian targets; Trajectory tracking: establish trajectory objects, determine the relationship between characters in each frame of the image, and form a tracking trajectory; Behavior judgment: Calculate the direction of the pedestrian target based on the tracking trajectory; obtain the time when the pedestrian target enters the zebra crossing and the time when the traffic light switches to red; determine whether the pedestrian runs a red light; Capture pictures: capture images of pedestrians running red lights and save them; Output result information: Get the time when the image of the pedestrian running a red light is taken, and output it together with the image.
[0022] In the target detection step, non-pedestrian targets usually include forklifts, bicycles, convertibles, etc. The people on these vehicles will also be captured by the camera. Therefore, when the video frame image is analyzed, the people on these vehicles will also be captured as human bounding boxes, but these people are not pedestrians, so they will be filtered out in the target filtering and extraction step, reducing the number of people whose illegal behaviors are finally analyzed and improving the accuracy of recognition.
[0023] In the target filtering and extraction step, the deep learning model method is used to identify objects, which can determine the target type very accurately. In addition, the specific operations of analyzing and filtering out non-human targets and non-pedestrian targets to extract pedestrian targets include: judging the human target frames contained in the non-human target frames or half-enclosed in the non-human frame according to the frame coordinates, and filtering out the frames of these human targets; filtering out all non-pedestrian target frames; and extracting the targets corresponding to the remaining frames as pedestrian targets.
[0024] In the trajectory tracking step, a trajectory object is established, and the character relationship of each frame image is determined. The specific operations of forming the tracking trajectory include: establishing trajectory points for pedestrian targets in the image; cyclically comparing the pedestrian targets in the current video frame image with the pedestrian targets in the previous video frame image according to the time sequence to determine whether the pedestrian targets appear for the first time; if it is not the first time, the trajectory points of the pedestrian targets in the two video frame images are associated to form the trajectory route of the pedestrian.
[0025] Among them, the specific contents of judging whether the pedestrian target appears for the first time include: cyclically comparing the degree of overlap between the pedestrian target in the current video frame image and the border of the pedestrian target in the previous video frame image; predicting the position of the pedestrian target based on the Kalman filter prediction method; setting a first threshold for the degree of overlap, if the degree of overlap is greater than or equal to the first threshold, and the difference between the predicted value and the actual value of the pedestrian target position is less than or equal to the second threshold, it is judged as the same pedestrian target that does not appear for the first time; otherwise, it is judged as the first appearance, determined as a new target, and its value is initialized.
[0026] The value range of the first threshold can be set to 0.75 to 0.85, usually 0.8; the value range of the second threshold can be set to 10 to 20 pixels, which can be adjusted according to the actual scene. In the above operation, the Kalman filter prediction method is combined to predict the target position. If the predicted value is significantly different from the actual value, even if the overlap is greater than or equal to the first threshold, it is not considered to be the same target. This can significantly reduce the misjudgment of identifying the borders of different pedestrians as belonging to the same pedestrian, thereby improving the accuracy of recognition.
[0027] In the behavior judgment step, the specific operation content of calculating the direction of travel of the pedestrian target according to the tracking trajectory includes: setting a third threshold value for the number of trajectory points, and when the number of trajectory points in the trajectory line is greater than or equal to the third threshold value, taking the extension direction of the trajectory line as the direction of travel of the pedestrian target. Obtaining the time when the pedestrian target enters the zebra crossing includes: obtaining the time of the video frame image captured when the pedestrian enters the zebra crossing (that is, the time when the trajectory line of the pedestrian target first enters the zebra crossing), and this time is used as the time when the pedestrian target enters the zebra crossing.
[0028] Among them, the value range of the third threshold is greater than 3. According to more than three trajectory points, the direction of travel of the pedestrian target can be judged more accurately, thereby improving the recognition accuracy.
[0029] In the behavior judgment step, the specific contents of judging whether a pedestrian runs a red light include: Determine whether the pedestrian target is moving in the direction of crossing the zebra crossing; if not, remove the pedestrian target; this operation can remove pedestrian targets that appear on the zebra crossing but do not run a red light, such as a cleaner moving perpendicular to the direction of running a red light; Determine whether the traffic light is green when the pedestrian enters the zebra crossing; if so, remove the pedestrian; in this operation, it is not necessary to set the 5-second red light transition time to improve the recognition accuracy; The remaining pedestrians that were not eliminated were targets of pedestrians running red lights.
[0030] Among them, at intersections with yellow zebra crossing lights, the yellow light will be treated as a green light or a red light according to local traffic regulations. For example, if it is stipulated that pedestrians who enter the zebra crossing area when the yellow light is on are in violation of the traffic rules, the yellow light will be treated as a red light.
[0031] Compared with the inherent defects of using time transition, which will lead to misjudgment and missed judgment, this patent solution uses a trajectory tracking method. As long as the time when the person enters the zebra crossing and the status of the traffic light at that time are determined, it can accurately determine whether the person is a person who runs a red light, thereby avoiding misjudgment and missed judgment; through the deep learning method to identify objects, the target type can be determined very accurately, so as to filter out people on forklifts and other objects, making the positioning of pedestrians more accurate; at the same time, considering the influence of other pedestrians in public places, the direction of travel is used to locate people who need to pass the traffic light, filtering out the interference of people with other behaviors on the capture. This method is very useful for units that need to manage, educate, and punish people who run red lights.
[0032] The target detection part of the above scheme can also use traditional image processing technology to achieve the target detection function without using deep learning models, such as: using color-based object judgment, edge detection algorithm-based shape recognition, and contour analysis-based shape matching. These technologies are mainly used to collect target borders. They do not need to collect faces in the early stage and compare them with the background database. They only need to obtain face data after accurately obtaining the target of pedestrians running red lights to compare with the background database, which can greatly reduce the amount of data analysis and make data acquisition more convenient.
[0033] The invention discloses a pedestrian red light violation recognition and capture system, which comprises a processor and a memory, wherein the memory stores a plurality of instructions; the processor loads instructions from the memory to execute the pedestrian red light violation recognition and capture method as described above.
[0034] The present invention discloses an electronic device, comprising a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the above-mentioned method for identifying and capturing pedestrians running red lights.
[0035] The present invention discloses a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for identifying and capturing pedestrians running red lights as described above are implemented.
[0036] Although the above methods are illustrated and described as a series of actions for simplicity of explanation, it should be understood and appreciated that these methods are not limited by the order of the actions, because according to one or more embodiments, some actions may occur in different orders and / or concurrently with other actions from the illustrations and descriptions herein or not illustrated and described herein but understandable to those skilled in the art. It will be further appreciated by those skilled in the art that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or a combination of the two. To clearly explain this interchangeability of hardware and software, various illustrative components, frames, modules, circuits, and steps are generally described in the form of their functionality above. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. The technical staff can implement the described functionality in different ways for each specific application, but such implementation decisions should not be interpreted as resulting in a departure from the scope of the present invention. The various illustrative logic blocks, modules, and circuits described in conjunction with the embodiments disclosed herein may be implemented or executed with a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor may be a microprocessor, but in an alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors cooperating with a DSP core, or any other such configuration. The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor so that the processor can read and write information from / to the storage medium. In an alternative, the storage medium may be integrated into the processor. The processor and storage medium may reside in an ASIC. The ASIC may reside in a user terminal. In an alternative, the processor and storage medium may reside in a user terminal as discrete components. In one or more exemplary embodiments, the functions described may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software as a computer program product, the functions may be stored or transmitted as one or more instructions or codes on a computer-readable medium.Computer-readable medium includes both computer storage medium and communication medium, and it includes any medium that facilitates computer program to transfer from one place to another place. Storage medium can be any available medium that can be accessed by computer. As an example and not limitation, such computer-readable medium can include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage or other magnetic storage device, or can be used to carry or store the desired program code of instruction or data structure form and any other medium that can be accessed by computer. Any connection is also properly referred to as computer-readable medium. For example, if software is transmitted from website, server or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL) or wireless technology such as infrared, radio and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL or wireless technology such as infrared, radio and microwave are just included in the definition of medium. Disk and disc as used herein include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and Blu-ray disc, wherein disk often reproduces data magnetically, while disc reproduces data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media.
[0037] The embodiments of the present invention are described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Various changes can be made within the knowledge scope of ordinary technicians in the technical field without departing from the purpose of the present invention.
Claims
1. A method for identifying and capturing pedestrians running red lights, characterized in that: include: Video stream data acquisition: collect video data and split the video into video frame images; Target detection: detect each frame of the image and obtain the bounding boxes, bounding box coordinates and time of human targets and non-human targets; The human targets include pedestrian targets and non-pedestrian targets; Target filtering and extraction: Analyze and filter out non-human targets and non-pedestrian targets to extract pedestrian targets; Trajectory tracking: establish trajectory objects, determine the relationship between characters in each frame of the image, and form a tracking trajectory; Behavior judgment: Calculate the direction of the pedestrian target based on the tracking trajectory; obtain the time when the pedestrian target enters the zebra crossing and the time when the traffic light switches to red; Determine whether a pedestrian runs a red light; Capture pictures: capture images of pedestrians running red lights and save them; Output result information: Get the time when the image of the pedestrian running a red light is taken, and output it together with the image.
2. The method for identifying and capturing pedestrians running red lights according to claim 1, characterized in that: In the target filtering and extraction step, non-human targets and non-pedestrian targets are analyzed and filtered out to extract pedestrian targets including: According to the frame coordinates, the frames of human targets that are contained in the non-human frame or half-enclosed in the non-human frame are determined, and the frames of these human targets are filtered out; Filter out all bounding boxes of non-pedestrian targets; The objects corresponding to the remaining bounding boxes are extracted as pedestrian objects.
3. The method for identifying and capturing pedestrians running red lights according to claim 2, characterized in that: In the trajectory tracking step, a trajectory object is established to determine the character relationship in each frame image. The tracking trajectory is formed including: Establish trajectory points for pedestrian targets in the image; According to the time sequence, the pedestrian target in the current video frame image is cyclically compared with the pedestrian target in the previous video frame image to determine whether the pedestrian target appears for the first time; if it is not the first time, the trajectory points of the pedestrian target in the two video frames are associated to form the pedestrian's trajectory line.
4. The method for identifying and capturing pedestrians running red lights according to claim 3, characterized in that: In the trajectory tracking step, the content of judging whether the pedestrian target appears for the first time includes: Cyclic comparison of the overlapping degree of the pedestrian target in the current video frame image and the border of the pedestrian target in the previous video frame image; predicting the position of the pedestrian target based on the Kalman filter prediction method; A first threshold is set for the degree of overlap. If the degree of overlap is greater than or equal to the first threshold, and the difference between the predicted value and the actual value of the pedestrian target position is less than or equal to the second threshold, it is judged as the same pedestrian target that does not appear for the first time; otherwise, it is judged as the first appearance.
5. The method for identifying and capturing pedestrians running red lights according to claim 4, characterized in that: In the behavior judgment step, Calculating the direction of travel of the pedestrian target according to the tracking trajectory includes: setting a third threshold for the number of trajectory points, and when the number of trajectory points in the trajectory line is greater than or equal to the third threshold, taking the extension direction of the trajectory line as the direction of travel of the pedestrian target; Acquiring the time when the pedestrian target enters the zebra crossing includes: acquiring the time of the video frame image captured when the pedestrian enters the zebra crossing, and the time is used as the time when the pedestrian target enters the zebra crossing.
6. The method for identifying and capturing pedestrians running red lights according to claim 5, characterized in that: 0.75≤first threshold≤0.85, 10 pixels≤second threshold≤20 pixels, 3≤third threshold.
7. The method for identifying and capturing pedestrians running red lights according to claim 5 or 6, characterized in that: In the behavior judgment step, judging whether a pedestrian runs a red light includes: Determine whether the pedestrian target is moving in the direction of crossing the zebra crossing; if not, remove the pedestrian target; Determine whether the traffic light is green when the pedestrian enters the zebra crossing; if so, remove the pedestrian; The remaining pedestrians that were not eliminated were targets of pedestrians running red lights.
8. The pedestrian red light running recognition and capture system is characterized by: It comprises a processor and a memory, wherein the memory stores a plurality of instructions; the processor loads instructions from the memory to execute the method for identifying and capturing pedestrians running red lights as described in any one of claims 1 to 7.
9. An electronic device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method for identifying and capturing pedestrians running red lights as described in any one of claims 1-7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for identifying and capturing pedestrians running red lights as described in any one of claims 1 to 7 are implemented.