Method, device and medium for target tracking of a robot

Abstract

This application discloses a method, device, and medium for target tracking in robots. The method includes: receiving real-time video captured by the robot's camera system; identifying an initial frame image using a pre-trained recognition model to determine a target to be tracked in the initial frame image; configuring a corresponding tracker for the target to be tracked; and inputting feature information into the tracker; acquiring the current frame image of the real-time video; and performing detection and tracking in the current frame image using the tracker and based on the feature information to obtain the tracking result of the target to be tracked. This improves the accuracy and stability of target tracking applied to robots, does not radiate radio waves during operation, is not easily detected by electronic reconnaissance equipment, and enhances stealth and anti-electronic interference capabilities.

Description

Technical Field

[0001] This application relates to the field of robot target recognition, specifically to a method, device, and medium for robot target tracking. Background Technology

[0002] With the continuous development of robot target recognition technology, robot target tracking has gradually become a hot topic. For example, in service robots, by recognizing and tracking human bodies or faces, it is possible to help humans determine the movement status of targets.

[0003] Currently, most robot tracking technologies rely on hardware solutions, such as radar or Bluetooth. However, hardware is susceptible to external factors like radio waves, leading to lower tracking accuracy and poorer tracking performance. Summary of the Invention

[0004] To address the aforementioned problems, namely the low tracking accuracy and poor tracking performance in hardware-based robot tracking technologies, this application proposes a method, device, and medium for robot target tracking, including:

[0005] This application proposes a method, apparatus, and medium for target tracking in a robot, including:

[0006] On one hand, this application provides a method for target tracking of a robot, comprising: receiving real-time video captured by a robot's camera system, wherein the real-time video contains a target to be tracked; acquiring an initial frame image of the real-time video, and recognizing the initial frame image using a pre-trained recognition model to determine the target to be tracked in the initial frame image, and feature information of the target to be tracked; configuring a corresponding tracker for the target to be tracked, and inputting the feature information into the tracker; acquiring a current frame image of the real-time video, and performing detection and tracking in the current frame image using the tracker and based on the feature information to obtain a tracking result for the target to be tracked.

[0007] In one example, the current frame image of the real-time video is acquired, and the tracker, along with the feature information, performs detection and tracking in the current frame image to obtain the tracking result of the target to be tracked. Specifically, this includes: acquiring the current frame image of the real-time video; obtaining multiple target recognitions through the tracker and the current frame image; obtaining the intersection-over-union ratio (IoU) of the multiple target recognitions with the feature information; using a pre-stored Hungarian algorithm, determining the target recognition with the highest IoU as the approximate position of the target to be tracked; acquiring the previous frame image of the real-time video and inputting the previous frame image into the Kalman filter in the tracker to obtain the predicted position of the target to be tracked, wherein the previous frame image is the image corresponding to the frame preceding the current frame image; and obtaining the tracking result of the target to be tracked based on the approximate position of the target to be tracked, the predicted position of the target to be tracked, and a pre-stored covariance prediction algorithm.

[0008] In one example, after obtaining the tracking result of the target based on the approximate location of the target, the predicted location of the target, and a pre-stored covariance prediction algorithm, the method further includes: calculating the Kalman gain data, position update data, and covariance update data of the tracker based on the tracking result of the target, the predicted location of the target, and the approximate location of the target; and updating the tracker based on the Kalman gain data, the position update data, and the covariance update data.

[0009] In one example, before determining the approximate location of the target to be tracked by using a pre-stored Hungarian algorithm to identify the target corresponding to the highest intersection-union ratio, the method further includes: determining, based on the intersection-union ratio, that the tracker does not have a target matching the feature information, so as to initialize the tracker.

[0010] In one example, after acquiring the current frame image of the real-time video and performing detection and tracking in the current frame image using the tracker and based on the feature information to obtain the tracking result of the target to be tracked, the method further includes: constructing a coordinate system based on the current frame image, with the robot's location as the starting point; determining the coordinates of the target to be tracked based on the position of the tracking result of the target to be tracked in the current frame image; determining the target coordinates of the robot based on the coordinates of the target to be tracked; and constructing a movement path for the robot based on the target coordinates.

[0011] In one example, determining the robot's target coordinates based on the coordinates of the target to be tracked specifically includes: determining the robot's tracking distance based on a pre-stored tracking strategy; determining the robot's undetermined coordinates based on the target's coordinates and the tracking distance; determining the presence of an obstacle at the undetermined coordinates based on the current frame image, and obtaining the appearance parameters of the obstacle; determining an offset value based on the appearance parameters and relevant parameters of the robot, and determining the robot's target coordinates based on the offset value and the undetermined coordinates.

[0012] In one example, after constructing a movement path for the robot based on the target coordinates, the method further includes: acquiring radar detection data in real time through the robot's radar system; determining local obstacle avoidance instructions based on the radar detection data; and sending the local obstacle avoidance instructions to the robot so that the robot moves to the target coordinates according to the movement path and the local obstacle avoidance instructions.

[0013] In one example, the method further includes: the robot has a built-in robot operating system; and establishing communication with the robot through the robot operating system.

[0014] On the other hand, this application also provides a target tracking device for a robot, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the following instructions: receiving real-time video captured by a robot's camera system, wherein the real-time video contains a target to be tracked; acquiring an initial frame image of the real-time video, and recognizing the initial frame image using a pre-trained recognition model to determine the target to be tracked in the initial frame image, and feature information of the target to be tracked; configuring a corresponding tracker for the target to be tracked, and inputting the feature information into the tracker; acquiring a current frame image of the real-time video, and performing detection and tracking in the current frame image using the tracker and based on the feature information to obtain a tracking result for the target to be tracked.

[0015] On the other hand, this application also provides a non-volatile computer storage medium storing computer-executable instructions, the computer-executable instructions being configured to: receive real-time video captured by a robot's camera system, wherein the real-time video contains a target to be tracked; acquire an initial frame image of the real-time video, and identify the initial frame image using a pre-trained recognition model to determine the target to be tracked in the initial frame image, and the feature information of the target to be tracked; configure a corresponding tracker for the target to be tracked, and input the feature information into the tracker; acquire the current frame image of the real-time video, and perform detection and tracking in the current frame image using the tracker and based on the feature information to obtain the tracking result of the target to be tracked.

[0016] The target tracking method, device, and medium for robots proposed in this application can bring the following beneficial effects: improve the accuracy and stability of target tracking applied to robots, do not radiate radio waves during operation, are not easily detected by electronic reconnaissance equipment, and improve stealth and anti-electronic interference capabilities. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0018] Figure 1 This is a flowchart illustrating a target tracking method for a robot according to an embodiment of this application;

[0019] Figure 2 This is a schematic diagram of a target tracking device for a robot according to an embodiment of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] First, it should be noted that the robot target tracking method described in this application can be set up in a corresponding system or server. This system or server establishes communication with the robot through a corresponding communication module to determine whether to receive information or send instructions. Furthermore, the terminal or server where the system resides should have built-in corresponding hardware devices, including but not limited to: processors, memory, communication modules, etc., to implement the various technical solutions in this application. In this application embodiment, a system is used as an example for explanation. This system can be set up in a corresponding terminal, including but not limited to: mobile phones, tablets, computers, or other terminal devices with corresponding computing power and functions. The system can determine its interaction relationship with the hardware devices in the terminal through corresponding program settings. Simultaneously, it can also determine the program's opening method through corresponding software settings, including but not limited to direct opening, opening via APP, WEB webpage login, etc., to meet the user's operation, monitoring, or debugging of the system, thereby achieving robot target tracking.

[0022] The technical solutions provided by the various embodiments of this application are described in detail below with reference to the accompanying drawings.

[0023] like Figure 1 As shown in the embodiment of this application, a method for target tracking of a robot includes:

[0024] S101: Receive real-time video captured by the robot's camera system, wherein the real-time video contains the target to be tracked.

[0025] Specifically, the robot is equipped with a camera system, which should include cameras capable of capturing high-definition images to meet preset requirements. In addition, the robot may be equipped with a power system to track and lock onto targets.

[0026] Furthermore, the camera system can monitor the robot's surrounding environment and generate corresponding real-time video. Based on the robot tracking problem that needs to be solved in this application, the real-time video contains the target to be tracked.

[0027] The system receives real-time video from the robot's camera system to track the target based on the real-time video.

[0028] S102: Acquire the initial frame image of the real-time video, and identify the initial frame image using a pre-trained recognition model to determine the target to be tracked in the initial frame image, as well as the feature information of the target to be tracked.

[0029] Specifically, before tracking the target, the system first needs to confirm the target. The system acquires the initial frame image of the real-time video, which is the first frame image of the real-time video, and the initial frame image contains the target to be tracked.

[0030] Furthermore, the system stores a pre-trained recognition model, which is trained based on the target to be tracked and can accurately identify the target based on its relevant features. The system uses the pre-trained recognition model to identify the initial frame image to determine the target to be tracked in the initial frame image and its feature information.

[0031] The recognition model obtains the target bounding box of the target to be tracked through a fully connected layer. The target bounding box is used to select the target to be tracked. The subsequent tracking process only needs to identify and track the target bounding box to complete the corresponding task.

[0032] In addition, it should be noted that the number of targets to be tracked is not limited in this application; it can be a single target or multiple targets.

[0033] S103: Configure a corresponding tracker for the target to be tracked, and input the feature information into the tracker.

[0034] Specifically, in order to track the target, a corresponding tracker needs to be configured for the target. If there are multiple targets, then a corresponding number of trackers should be configured.

[0035] In addition, to ensure that the tracker can accurately identify and track the target, the system needs to input feature information into the tracker so that the tracker can accurately identify and track the target based on the feature information.

[0036] S104: Obtain the current frame image of the real-time video, and perform detection and tracking in the current frame image using the tracker and based on the feature information to obtain the tracking result of the target to be tracked.

[0037] Specifically, the system acquires the current frame image of a real-time video, and uses the tracker and the current frame image to obtain multiple recognition targets. These multiple recognition targets refer to the results of identifying and selecting all objects or human bodies appearing in the current frame image. The multiple recognition targets may include the target to be tracked.

[0038] To determine the target to be tracked among multiple identification targets, it is also necessary to obtain the Intersection-Over-Union (IOU) ratio between the multiple identification targets and the feature information. Here, IOU is determined by comparing the similarity with the feature information. If the similarity is high, the IOU is high; if the similarity is low, the IOU is low.

[0039] The system uses a pre-stored Hungarian algorithm to identify the target with the highest Intersection over Union (IOU) and determines the approximate location of the target to be tracked. It should be noted that the approximate location here is in the form of a bounding box.

[0040] Furthermore, to ensure tracking accuracy, this application introduces Kalman filtering technology. The tracker includes a Kalman filter. By acquiring the previous frame of the real-time video and inputting it into the Kalman filter in the tracker, the system can obtain the predicted position of the target to be tracked. It should be noted that the previous frame here refers to the image corresponding to the frame preceding the current frame.

[0041] The Kalman filter can infer the movement trajectory of the target to be tracked based on the previous frame image, and thus obtain the predicted position of the target to be tracked. Similarly, the predicted position here is also in the form of a target box.

[0042] Based on the approximate location of the target, the predicted location of the target, and the pre-stored covariance prediction algorithm, the tracking result of the target can be obtained. This tracking result represents the most reasonable and accurate current location of the target.

[0043] In one embodiment, after obtaining the tracking result of the target based on the approximate location of the target, the predicted location of the target, and the pre-stored covariance prediction algorithm, the system can also update the tracker to improve the final accuracy.

[0044] Specifically, the system calculates the tracker's Kalman gain data, position update data, and covariance update data based on the tracking results of the target, the predicted position of the target, and the approximate position of the target.

[0045] The tracker can be updated based on Kalman gain data, position update data, and covariance update data, ensuring that the tracker can more accurately match the target to be tracked.

[0046] In one embodiment, before the system determines the target corresponding to the highest Intersection over Union (IOU) as the approximate location to be tracked using the Hungarian algorithm, it can also determine, based on the IOU, that the tracker does not have a target matching the feature information. In this case, it can be determined that the target to be identified is not included among multiple targets. If such a situation exists, the tracker is initialized.

[0047] In one embodiment, the system acquires the current frame image of a real-time video, and performs detection and tracking in the current frame image using a tracker and based on feature information. After obtaining the tracking result of the target to be tracked, the system can also control the robot to move in order to achieve the tracking of the target to be tracked.

[0048] Specifically, the system constructs a coordinate system based on the current frame image, with the robot's location as the starting point.

[0049] The system determines the coordinates of the target by the position of the target in the current frame image based on the tracking results and by combining the coordinate system.

[0050] Furthermore, the system determines the robot's target coordinates based on the coordinates of the target to be tracked. It should be noted that because the tracking process requires maintaining a certain distance from the target, the target coordinates are not the same as the coordinates of the target being tracked.

[0051] Furthermore, the system can construct a movement path for the robot based on the target coordinates, so that the robot can move to the target coordinates according to the movement path.

[0052] In one embodiment, the system determines the robot's target coordinates based on the coordinates of the target to be tracked, specifically including:

[0053] The system determines the robot's tracking distance based on the pre-stored tracking strategy. Here, the tracking distance is the distance between the robot's target coordinates and the coordinates of the target to be tracked.

[0054] Furthermore, the tracking strategy can be configured to adjust the tracking distance based on the robot's surrounding environment. For example, when there are many targets to identify, the tracking distance threshold can be lowered, and when there are fewer targets to identify, the tracking distance threshold can be increased accordingly.

[0055] Furthermore, the system can determine the robot's undetermined coordinates based on the coordinates of the target being tracked and the tracking distance. These undetermined coordinates are the target coordinates if there are no obstacles there, and the robot can move to these coordinates.

[0056] However, real-world environments are often very complex. Therefore, the system also needs to determine whether an obstacle exists at the desired coordinates based on the current frame image. If it is determined that an obstacle exists, the system obtains the obstacle's appearance parameters. These appearance parameters can include the obstacle's width, height, and other information.

[0057] Furthermore, the system determines the offset value based on the appearance parameters and relevant robot parameters, which may include the robot's width and height. To avoid collisions between the robot and obstacles, the offset value needs to be determined based on the aforementioned appearance and relevant parameters. This offset value is the distance from the predetermined coordinates that allows the robot to avoid obstacles.

[0058] Furthermore, the system determines the robot's target coordinates based on the offset value and the coordinates to be determined.

[0059] In one embodiment, after the system constructs a movement path for the robot based on the target coordinates, an obstacle avoidance scheme is also involved to ensure that the robot can avoid obstacles during its movement.

[0060] Specifically, the system acquires radar detection data in real time through the robot's radar system and determines local obstacle avoidance commands based on the radar detection data. These local obstacle avoidance commands are commands to control the robot to stop, turn, or move straight.

[0061] Furthermore, the system sends local obstacle avoidance commands to the robot, enabling the robot to move to the target coordinates according to the movement path and the local obstacle avoidance commands.

[0062] In one embodiment, the robot has a built-in Robot Operating System (ROS). The system can communicate with the robot through the ROS to receive information from the robot and / or send instructions to the robot.

[0063] In one embodiment, such as Figure 2 As shown, this application also provides a device for target tracking of a robot, comprising:

[0064] At least one processor; and,

[0065] A memory communicatively connected to the at least one processor; wherein,

[0066] The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to perform the following instructions:

[0067] Receive real-time video captured by the robot's camera system, wherein the real-time video contains the target to be tracked;

[0068] The initial frame image of the real-time video is acquired, and the initial frame image is identified by a pre-trained recognition model to determine the target to be tracked in the initial frame image and the feature information of the target to be tracked.

[0069] Configure a corresponding tracker for the target to be tracked, and input the feature information into the tracker;

[0070] The current frame image of the real-time video is acquired, and the tracker, based on the feature information, performs detection and tracking in the current frame image to obtain the tracking result of the target to be tracked.

[0071] In one embodiment, this application also provides a non-volatile computer storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured as follows:

[0072] Receive real-time video captured by the robot's camera system, wherein the real-time video contains the target to be tracked;

[0073] The initial frame image of the real-time video is acquired, and the initial frame image is identified by a pre-trained recognition model to determine the target to be tracked in the initial frame image and the feature information of the target to be tracked.

[0074] Configure a corresponding tracker for the target to be tracked, and input the feature information into the tracker;

[0075] The current frame image of the real-time video is acquired, and the tracker, based on the feature information, performs detection and tracking in the current frame image to obtain the tracking result of the target to be tracked.

[0076] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device and medium embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the description of the method embodiments.

[0077] The devices and media provided in this application are one-to-one with the methods. Therefore, the devices and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0078] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0079] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0080] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0081] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0082] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0083] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0084] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, 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 technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0085] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0086] The foregoing is merely an 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 changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A method for target tracking in a robot, characterized in that, include: Receive real-time video captured by the robot's camera system, wherein the real-time video contains the target to be tracked; The initial frame image of the real-time video is acquired, and the initial frame image is identified by a pre-trained recognition model to determine the target to be tracked in the initial frame image and the feature information of the target to be tracked. Configure a corresponding tracker for the target to be tracked, and input the feature information into the tracker; The current frame image of the real-time video is acquired, and the tracker, based on the feature information, performs detection and tracking in the current frame image to obtain the tracking result of the target to be tracked. The process involves acquiring the current frame image of the real-time video, and using the tracker and based on the feature information to perform detection and tracking in the current frame image to obtain the tracking result of the target to be tracked. Specifically, this includes: The current frame image of the real-time video is acquired, and multiple recognition targets are obtained through the tracker and the current frame image, and the intersection-union ratio of the multiple recognition targets and the feature information is obtained; Using the pre-stored Hungarian algorithm, the target corresponding to the highest intersection-union ratio is determined as the approximate location of the target to be tracked; The previous frame image of the real-time video is acquired and input into the Kalman filter in the tracker to obtain the predicted position of the target to be tracked, wherein the previous frame image is the image corresponding to the previous frame of the current frame image; Based on the approximate location of the target to be tracked, the predicted location of the target to be tracked, and the pre-stored covariance prediction algorithm, the tracking result of the target to be tracked is obtained. After obtaining the tracking result of the target based on the approximate location of the target, the predicted location of the target, and a pre-stored covariance prediction algorithm, the method further includes: Based on the tracking results of the target to be tracked, the predicted position of the target to be tracked, and the approximate position of the target to be tracked, the Kalman gain data, position update data, and covariance update data of the tracker are calculated. The tracker is updated based on the Kalman gain data, the position update data, and the covariance update data; After acquiring the current frame image of the real-time video, and performing detection and tracking in the current frame image using the tracker and based on the feature information to obtain the tracking result of the target to be tracked, the method further includes: Based on the current frame image, a coordinate system is constructed with the robot's location as the starting point; The coordinates of the target to be tracked are determined based on the position of the target in the current frame image according to the tracking result of the target to be tracked; Based on the coordinates of the target to be tracked, determine the target coordinates of the robot; Based on the target coordinates, a movement path is constructed for the robot; Based on the coordinates of the target to be tracked, the target coordinates of the robot are determined, specifically including: The tracking distance of the robot is determined according to the pre-stored tracking strategy; The robot's undetermined coordinates are determined based on the coordinates of the target being tracked and the tracking distance. Based on the current frame image, it is determined that an obstacle exists at the unknown coordinates, and the appearance parameters of the obstacle are obtained; Based on the appearance parameters and the relevant parameters of the robot, the offset value is determined, and based on the offset value and the undetermined coordinates, the target coordinates of the robot are determined. After constructing a movement path for the robot based on the target coordinates, the method further includes: The robot's radar system acquires radar detection data in real time. Local obstacle avoidance commands are determined based on the radar detection data; The local obstacle avoidance command is sent to the robot so that the robot moves to the target coordinates according to the movement path and the local obstacle avoidance command.

2. The target tracking method for a robot according to claim 1, characterized in that, Before determining the approximate location of the target to be tracked by using a pre-stored Hungarian algorithm to identify the target corresponding to the highest intersection-union ratio, the method further includes: Based on the intersection-union ratio, it is determined that the tracker does not have a target that matches the feature information, so that the tracker is initialized.

3. The target tracking method for a robot according to claim 1, characterized in that, The method further includes: the robot has a built-in robot operating system; Communication is established with the robot through the robot operating system.

4. A device for target tracking of a robot, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to perform the following instructions: Receive real-time video captured by the robot's camera system, wherein the real-time video contains the target to be tracked; The initial frame image of the real-time video is acquired, and the initial frame image is identified by a pre-trained recognition model to determine the target to be tracked in the initial frame image and the feature information of the target to be tracked. Configure a corresponding tracker for the target to be tracked, and input the feature information into the tracker; The current frame image of the real-time video is acquired, and the tracker, based on the feature information, performs detection and tracking in the current frame image to obtain the tracking result of the target to be tracked. The process involves acquiring the current frame image of the real-time video, and using the tracker and based on the feature information to perform detection and tracking in the current frame image to obtain the tracking result of the target to be tracked. Specifically, this includes: The current frame image of the real-time video is acquired, and multiple recognition targets are obtained through the tracker and the current frame image, and the intersection-union ratio of the multiple recognition targets and the feature information is obtained; Using the pre-stored Hungarian algorithm, the target corresponding to the highest intersection-union ratio is determined as the approximate location of the target to be tracked; The previous frame image of the real-time video is acquired and input into the Kalman filter in the tracker to obtain the predicted position of the target to be tracked, wherein the previous frame image is the image corresponding to the previous frame of the current frame image; Based on the approximate location of the target to be tracked, the predicted location of the target to be tracked, and the pre-stored covariance prediction algorithm, the tracking result of the target to be tracked is obtained. After obtaining the tracking result of the target based on the approximate location of the target, the predicted location of the target, and the pre-stored covariance prediction algorithm, the process further includes: Based on the tracking results of the target to be tracked, the predicted position of the target to be tracked, and the approximate position of the target to be tracked, the Kalman gain data, position update data, and covariance update data of the tracker are calculated. The tracker is updated based on the Kalman gain data, the position update data, and the covariance update data; After acquiring the current frame image of the real-time video, and performing detection and tracking in the current frame image using the tracker and based on the feature information to obtain the tracking result of the target to be tracked, the method further includes: Based on the current frame image, a coordinate system is constructed with the robot's location as the starting point; The coordinates of the target to be tracked are determined based on the position of the target in the current frame image according to the tracking result of the target to be tracked; Based on the coordinates of the target to be tracked, determine the target coordinates of the robot; Based on the target coordinates, a movement path is constructed for the robot; Based on the coordinates of the target to be tracked, the target coordinates of the robot are determined, specifically including: The tracking distance of the robot is determined according to the pre-stored tracking strategy; The robot's undetermined coordinates are determined based on the coordinates of the target being tracked and the tracking distance. Based on the current frame image, it is determined that an obstacle exists at the unknown coordinates, and the appearance parameters of the obstacle are obtained; Based on the appearance parameters and the relevant parameters of the robot, the offset value is determined, and based on the offset value and the undetermined coordinates, the target coordinates of the robot are determined. After constructing a movement path for the robot based on the target coordinates, the process further includes: The robot's radar system acquires radar detection data in real time. Local obstacle avoidance commands are determined based on the radar detection data; The local obstacle avoidance command is sent to the robot so that the robot moves to the target coordinates according to the movement path and the local obstacle avoidance command.

5. A non-volatile computer storage medium storing computer-executable instructions, characterized in that, The computer-executable instructions are set as follows: Receive real-time video captured by the robot's camera system, wherein the real-time video contains the target to be tracked; The initial frame image of the real-time video is acquired, and the initial frame image is identified by a pre-trained recognition model to determine the target to be tracked in the initial frame image and the feature information of the target to be tracked. Configure a corresponding tracker for the target to be tracked, and input the feature information into the tracker; The current frame image of the real-time video is acquired, and the tracker, based on the feature information, performs detection and tracking in the current frame image to obtain the tracking result of the target to be tracked. The process involves acquiring the current frame image of the real-time video, and using the tracker and based on the feature information to perform detection and tracking in the current frame image to obtain the tracking result of the target to be tracked. Specifically, this includes: The current frame image of the real-time video is acquired, and multiple recognition targets are obtained through the tracker and the current frame image, and the intersection-union ratio of the multiple recognition targets and the feature information is obtained; Using the pre-stored Hungarian algorithm, the target corresponding to the highest intersection-union ratio is determined as the approximate location of the target to be tracked; The previous frame image of the real-time video is acquired and input into the Kalman filter in the tracker to obtain the predicted position of the target to be tracked, wherein the previous frame image is the image corresponding to the previous frame of the current frame image; Based on the approximate location of the target to be tracked, the predicted location of the target to be tracked, and the pre-stored covariance prediction algorithm, the tracking result of the target to be tracked is obtained. After obtaining the tracking result of the target based on the approximate location of the target, the predicted location of the target, and the pre-stored covariance prediction algorithm, the process further includes: Based on the tracking results of the target to be tracked, the predicted position of the target to be tracked, and the approximate position of the target to be tracked, the Kalman gain data, position update data, and covariance update data of the tracker are calculated. The tracker is updated based on the Kalman gain data, the position update data, and the covariance update data; After acquiring the current frame image of the real-time video, and performing detection and tracking in the current frame image using the tracker and based on the feature information to obtain the tracking result of the target to be tracked, the method further includes: Based on the current frame image, a coordinate system is constructed with the robot's location as the starting point; The coordinates of the target to be tracked are determined based on the position of the target in the current frame image according to the tracking result of the target to be tracked; Based on the coordinates of the target to be tracked, determine the target coordinates of the robot; Based on the target coordinates, a movement path is constructed for the robot; Based on the coordinates of the target to be tracked, the target coordinates of the robot are determined, specifically including: The tracking distance of the robot is determined according to the pre-stored tracking strategy; The robot's undetermined coordinates are determined based on the coordinates of the target being tracked and the tracking distance. Based on the current frame image, it is determined that an obstacle exists at the unknown coordinates, and the appearance parameters of the obstacle are obtained; Based on the appearance parameters and the relevant parameters of the robot, the offset value is determined, and based on the offset value and the undetermined coordinates, the target coordinates of the robot are determined. After constructing a movement path for the robot based on the target coordinates, the process further includes: The robot's radar system acquires radar detection data in real time. Local obstacle avoidance commands are determined based on the radar detection data; The local obstacle avoidance command is sent to the robot so that the robot moves to the target coordinates according to the movement path and the local obstacle avoidance command.

Images