A target tracking method, device and computer readable storage medium

Abstract

The application discloses a target tracking method, which comprises the following steps: collecting image data and determining a tracking target in the image data; performing strong tracking matching on the tracking target and other targets according to a preset target strong tracking method; if the strong tracking matching of the tracking target is unsuccessful, performing conditional division on the tracking target, and performing weak tracking matching on the tracking target and other unmatched targets through a preset target weak tracking method according to the result of the conditional division; if the weak tracking matching of the tracking target is unsuccessful, obtaining a trajectory parameter of the tracking target, calculating a matching value of the tracking target and other unmatched targets according to the trajectory parameter, and performing supplementary matching on the trajectory of the tracking target according to the matching value. Thus, through three-level cascade matching of strong tracking, weak tracking and parameter matching on the tracking target, target matching can be performed when the target is tracked, and the problem that the target is easily lost when being occluded for a long time is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of target tracking, in particular to a target tracking method, device and computer readable storage medium. BACKGROUND

[0002] At present, the tracking method based on multiple targets has a wide range of applications in both military and civilian aspects, which also promotes the development of the multiple target tracking technology. However, there are still many difficulties in practical application engineering, such as the problem of missing detection or false detection of small targets in complex scenes, the problem of easy loss of large targets in the case of low frame rate, and the problem of easy loss of targets in the case of long-time occlusion. SUMMARY

[0003] The purpose of the present application is to provide a target tracking method, device and computer readable storage medium, which can help to match targets when tracking targets by performing three-level cascade matching of strong tracking, weak tracking and parameter matching, and solve the problem of easy loss of targets in the case of long-time occlusion.

[0004] To achieve the above purpose:

[0005] In a first aspect, the embodiments of the present application provide a target tracking method, which comprises the following steps:

[0006] Collecting image data and determining a tracking target in the image data;

[0007] Performing strong tracking matching of the tracking target and other targets according to a preset target strong tracking method, and if the strong tracking matching of the tracking target is successful, obtaining a behavior trajectory of the tracking target;

[0008] If the strong tracking matching of the tracking target is not successful, performing condition division on the tracking target, and according to the result of condition division, performing weak tracking matching of the tracking target and other unmatched targets by a preset target weak tracking method; if the weak tracking matching of the tracking target is successful, obtaining the behavior trajectory of the tracking target;

[0009] If the weak tracking matching of the tracking target is not successful, obtaining a trajectory parameter of the tracking target, and calculating a matching value of the tracking target and other unmatched targets according to the trajectory parameter, and performing supplementary matching of the trajectory of the tracking target according to the matching value.

[0010] Optionally, the collecting image data and determining a tracking target in the image data comprises:

[0011] Collecting image data information, performing target recognition and extraction on the image data information to obtain at least one candidate target;

[0012] According to area sizes of all or part of the alternative targets, the all or part of the alternative targets are classified into multiple grades, and target type identification is performed on the alternative targets of each grade;

[0013] According to the target type identification result, it is determined whether the alternative target of each grade is a tracking target.

[0014] Optionally, the target identification extraction on the image data information is performed to obtain at least one alternative target, and specifically includes:

[0015] At least one initial target is obtained through the target identification extraction;

[0016] It is judged whether a percentage of the number of each initial target in the number of all initial targets is greater than a preset proportion threshold value;

[0017] If yes, the initial target is determined as an alternative target;

[0018] If no, the initial target is determined as not an alternative target.

[0019] Optionally, the determination of whether the alternative target of each grade is a tracking target according to the target type identification result further includes:

[0020] The type of the target of each grade in the classification is identified, it is judged whether the target is misidentified according to the actual object of the target, a detection error rate of the type misidentification of the target is obtained, and it is judged whether the detection error rate is greater than a preset error threshold value;

[0021] If it is judged that the target detection error rate is greater than the preset error threshold value, a target corresponding to a highest error rate value in the target is determined as a tracking target;

[0022] If it is judged that the target detection error rate is less than or equal to the preset error threshold value, the type of the target is re-identified.

[0023] Optionally, if the strong tracking method matching of the tracking target is unsuccessful, the tracking target is conditionally divided, and according to a result of the conditional division, the tracking target and other unmatched targets are weakly tracked and matched through a preset target weak tracking method, including:

[0024] If the tracking target is not successfully matched with other unmatched targets through the target strong tracking, the area of the tracking target is compared with a preset area reference value;

[0025] if the area size of the tracking target exceeds the preset area reference value, the tracking target is divided into a large target, and a trajectory of the tracking target is tracked and matched based on a target feature comparison method;

[0026] if the area size of the tracking target does not exceed the preset area reference value, the tracking target is divided into a small target, and a trajectory of the tracking target is tracked and matched based on a filter tracking method.

[0027] Optionally, if the tracking target is not successfully matched by the weak tracking method, a trajectory parameter of the tracking target is acquired, and a matching value of the tracking target and other unmatched targets is calculated according to the trajectory parameter, including:

[0028] if the tracking target is not successfully matched by the weak tracking method, a trajectory parameter of the tracking target is acquired, and a matching value of the tracking target and other unmatched targets is calculated according to the trajectory parameter, including:

[0029] P = a1 * S + a2 * L

[0030] wherein, P represents the matching value of the tracking target, S represents the similarity of the tracking target, L represents the Euclidean distance between the tracking target and other matched targets, and a1 and a2 represent preset weighting parameters.

[0031] Optionally, the supplementary matching of the trajectory of the tracking target according to the matching value includes:

[0032] the matching value of the tracking target and other unmatched targets is calculated respectively, the maximum value of the calculated matching value of the tracking target is acquired, and it is judged whether the matching value is greater than a preset matching threshold value;

[0033] if it is judged that the matching value is greater than the preset matching threshold value, it is determined that the matching target corresponding to the matching value is successfully matched with the tracking target;

[0034] if it is judged that the matching value is less than or equal to the preset matching threshold value, it is determined that the tracking target is not successfully matched, and the tracking target is regarded as another new target.

[0035] In a second aspect, an embodiment of the present application provides a target tracking device, and the device includes:

[0036] a data acquisition module, configured to acquire image data and determine a tracking target in the image data;

[0037] a first-level tracking module configured to perform strong tracking matching on the tracking target and other targets according to a preset target strong tracking method, and if the strong tracking matching on the tracking target is successful, obtain a behavior trajectory of the tracking target;

[0038] a second-level tracking module configured to, if the strong tracking matching on the tracking target is unsuccessful, perform condition division on the tracking target, and according to a result of the condition division, perform weak tracking matching on the tracking target and other unmatched targets according to a preset target weak tracking method, and if the weak tracking matching on the tracking target is successful, obtain the behavior trajectory of the tracking target;

[0039] a third-level tracking module configured to, if the weak tracking matching on the tracking target is unsuccessful, obtain a trajectory parameter of the tracking target, and according to the trajectory parameter, calculate a matching value of the tracking target and other unmatched targets, and perform supplementary matching on the trajectory of the tracking target according to the matching value.

[0040] In a third aspect, an electronic device is disclosed, including: a memory storing executable program codes; and a processor coupled with the memory; the processor invokes the executable program codes stored in the memory, and is configured to execute the target tracking method according to the first aspect.

[0041] In a fourth aspect, a computer readable storage medium is provided, and the computer readable storage medium stores a computer program; when instructions in the computer readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to implement the target tracking method according to the first aspect.

[0042] The target tracking method provided by the embodiments of the present application includes the following steps: collecting image data, and determining a tracking target in the image data; performing strong tracking matching on the tracking target and other targets according to a preset target strong tracking method; if the strong tracking matching on the tracking target is unsuccessful, performing condition division on the tracking target, and according to a result of the condition division, performing weak tracking matching on the tracking target and other unmatched targets according to a preset target weak tracking method; if the weak tracking matching on the tracking target is unsuccessful, obtaining a trajectory parameter of the tracking target, and according to the trajectory parameter, calculating a matching value of the tracking target and other unmatched targets, and performing supplementary matching on the trajectory of the tracking target according to the matching value. In this way, by performing three-level cascade matching of strong tracking, weak tracking and parameter matching on the tracking target, target matching can be performed when tracking the target, and the problem that the target is easily lost due to long-time occlusion is solved. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1This is a flowchart illustrating a target tracking method according to a preferred embodiment of the present invention.

[0044] Figure 2 This is a schematic diagram of the structure of a target tracking device according to a preferred embodiment of the present invention;

[0045] Figure 3 This is a schematic diagram of a target tracking device provided in another preferred embodiment of the present invention. Detailed Implementation

[0046] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0047] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0048] See Figure 1 This application provides a target tracking method, which can be executed by a target tracking device provided in this application. The target tracking device can be implemented in software and / or hardware. In this embodiment, the target tracking device is applied to a server as an example. The target tracking method provided in this embodiment includes the following steps:

[0049] Step S101: Acquire image data and determine the tracking target in the image data.

[0050] In one embodiment, image data information is acquired, and target recognition and extraction are performed on the image data information to obtain at least one initial target. The initial target is statistically analyzed to obtain the percentage of the number of the initial target relative to the total number of identified initial targets. It is then determined whether the percentage of the initial target in the total number of initial targets is greater than a preset percentage threshold. If the percentage of the initial target in the total number of initial targets is greater than the preset percentage threshold, the initial target is determined to be a candidate target. If the percentage of the initial target in the total number of initial targets is less than or equal to the preset percentage threshold, the initial target is determined not to be a candidate target.

[0051] In one embodiment, at least one candidate target is obtained, and the candidate targets are classified into multiple levels based on their area size. The YOLOv5 algorithm is used to identify the target type and compare information for each level of candidate targets. The type of the target for each level of the classification is identified, and the target is judged to be incorrectly identified based on its actual object. The detection error rate of the incorrect target type identification is obtained, and it is determined whether the detection error rate is greater than a preset error threshold. If the target detection error rate is greater than the preset error threshold, the target with the highest error rate is identified as the tracking target. If the target detection error rate is less than or equal to the preset error threshold, the target type is re-identified.

[0052] Step S102: Perform strong tracking matching between the tracked target and other targets according to the preset target strong tracking method. If the strong tracking matching of the tracked target is successful, obtain the behavior trajectory of the tracked target.

[0053] In one embodiment, the target is subjected to strong tracking and matching with other targets according to a preset strong target tracking method. This strong target tracking method includes, but is not limited to, deep learning-based Bytetrack and Hungarian matching algorithms. Here, the strong target tracking method can adjust the tracking parameters for target tracking and matching based on different targets being tracked and matched, to ensure the accuracy of the current tracking effect.

[0054] Here, if the target being tracked is successfully matched with other targets using the target strong tracking method, the target strong tracking method will continue to be used to track and match the target.

[0055] Step S103: If the strong tracking match of the tracked target is unsuccessful, the tracked target is conditionally divided, and according to the result of the conditional division, the tracked target is weakly tracked and matched with other unmatched targets using a preset weak tracking method; if the weak tracking match of the tracked target is successful, the behavioral trajectory of the tracked target is obtained.

[0056] In one embodiment, if strong tracking matching of the tracked target fails, the target's area size is compared with a preset area benchmark value to perform conditional classification. Here, if the target's area size exceeds the preset area benchmark value, the target is classified as a large target, and the target's trajectory is tracked and matched based on a target feature comparison method. If the target's area size does not exceed the preset area benchmark value, the target is classified as a small target, and the target's trajectory is tracked and matched based on a filter-based tracking method.

[0057] Here, if the target being tracked is successfully matched with other targets using the target weak tracking method, the target weak tracking method will continue to be used to track and match the target.

[0058] Step S104: If the weak tracking match of the tracked target is unsuccessful, the trajectory parameters of the tracked target are obtained, and the matching value of the tracked target with other unmatched targets is calculated based on the trajectory parameters. The trajectory of the tracked target is then supplemented with matching based on the matching value.

[0059] In one embodiment, if the tracked target fails to match other unmatched targets through weak tracking, feature extraction is performed on the tracked target, and feature extraction is performed on the target corresponding to each trajectory to determine the similarity S of the tracked target; the Euclidean distance L between the center point of the tracked target and the center points of other unmatched targets in a frame of image data is calculated; and the matching value of the tracked target is calculated based on the similarity S and the Euclidean distance L.

[0060] P = a1*S + a2*L

[0061] Wherein, P represents the matching value of the tracked target, S represents the similarity of the tracked target, L represents the Euclidean distance between the tracked target and other matching targets, and a1 and a2 represent preset weighting parameters.

[0062] In one embodiment, the matching values ​​of the tracked target and other unmatched targets are calculated respectively, the maximum value among the calculated matching values ​​of the tracked target is obtained, and it is determined whether the matching value is greater than a preset matching threshold. Here, if it is determined that the matching value is greater than the preset matching threshold, it is determined that the matching target corresponding to the matching value is successfully matched with the tracked target. If it is determined that the matching value is less than or equal to the preset matching threshold, it is determined that the tracking target is unsuccessfully matched, and the tracked target is used as another new target.

[0063] In summary, the target tracking method provided in the above embodiments obtains the target through image data information and performs multi-target tracking by a three-level cascaded tracking matching based on strong target tracking, weak target tracking, and matching parameters. This helps to achieve a higher multi-target matching rate and better realize multi-target tracking.

[0064] Based on the same inventive concept as the foregoing embodiments, the method provided by the foregoing embodiments will be described in detail below through a specific example.

[0065] See Figure 2 The target tracking device provided in the application embodiment includes a data acquisition device, a primary tracking module, a secondary tracking module, and a tertiary tracking module, wherein...

[0066] The data acquisition module is used to acquire image data and determine the tracking target from the image data;

[0067] The first-level tracking module is used to perform strong tracking matching between the tracked target and other targets according to a preset strong tracking method. If the strong tracking matching of the tracked target is successful, the behavioral trajectory of the tracked target is obtained.

[0068] The secondary tracking module is used to perform conditional classification on the tracking target if strong tracking matching of the tracking target fails, and to perform weak tracking matching of the tracking target with other unmatched targets according to the result of conditional classification using a preset weak tracking method; if the weak tracking matching of the tracking target is successful, the behavioral trajectory of the tracking target is obtained.

[0069] The three-level tracking module is used to obtain the trajectory parameters of the tracking target if the weak tracking match of the tracking target fails, calculate the matching value of the tracking target with other unmatched targets based on the trajectory parameters, and perform supplementary matching of the trajectory of the tracking target based on the matching value.

[0070] In one embodiment, the data acquisition module is specifically used to acquire image data information, perform target recognition and extraction on the image data information to obtain at least one initial target, and statistically obtain the percentage of the number of the initial targets relative to the total number of identified initial targets, and determine whether the percentage of the initial targets in the total number of initial targets is greater than a preset percentage threshold. Here, if the percentage of the initial targets in the total number of initial targets is greater than the preset percentage threshold, the initial target is determined to be a candidate target; if the percentage of the initial targets in the total number of initial targets is less than or equal to the preset percentage threshold, the initial target is determined not to be a candidate target.

[0071] In one embodiment, at least one candidate target is obtained, and the candidate targets are classified into multiple levels based on their area size. The YOLOv5 algorithm is used to identify the target type and compare information for each level of candidate targets. The type of the target for each level of the classification is identified, and the target is judged to be incorrectly identified based on its actual object. The detection error rate of the incorrect target type identification is obtained, and it is determined whether the detection error rate is greater than a preset error threshold. If the target detection error rate is greater than the preset error threshold, the target with the highest error rate is identified as the tracking target. If the target detection error rate is less than or equal to the preset error threshold, the target type is re-identified.

[0072] In one embodiment, the primary tracking module is specifically used to perform strong tracking matching of the tracked target with other targets according to a preset strong target tracking method. The strong target tracking method includes, but is not limited to, deep learning-based Bytetrack, Hungarian matching algorithms, etc. Here, the strong target tracking method can adjust the tracking parameters for target tracking matching according to different tracked and matched targets to ensure the accuracy of the current tracking effect.

[0073] Here, if the target being tracked is successfully matched with other targets using the target strong tracking method, the target strong tracking method will continue to be used to track and match the target.

[0074] In one embodiment, the secondary tracking module is specifically used to, if the strong tracking match of the tracking target fails, divide the tracking target into conditions, and according to the result of the condition division, perform weak tracking match between the tracking target and other unmatched targets using a preset weak tracking method; if the weak tracking match of the tracking target succeeds, obtain the behavioral trajectory of the tracking target.

[0075] In one embodiment, if strong tracking matching of the tracked target fails, the target's area size is compared with a preset area benchmark value to perform conditional classification. Here, if the target's area size exceeds the preset area benchmark value, the target is classified as a large target, and the target's trajectory is tracked and matched based on a target feature comparison method. If the target's area size does not exceed the preset area benchmark value, the target is classified as a small target, and the target's trajectory is tracked and matched based on a filter-based tracking method.

[0076] Here, if the target being tracked is successfully matched with other targets using the target weak tracking method, the target weak tracking method will continue to be used to track and match the target.

[0077] In one embodiment, the three-level tracking module is specifically used to: if the tracked target fails to match other unmatched targets through weak target tracking, extract features from the tracked target and extract features from the target corresponding to each trajectory to determine the similarity S of the tracked target; calculate the Euclidean distance L between the center point of the tracked target and the center points of other unmatched targets in a frame of image data; and calculate the matching value of the tracked target based on the similarity S and the Euclidean distance L.

[0078] P = a1*S + a2*L

[0079] Wherein, P represents the matching value of the tracked target, S represents the similarity of the tracked target, L represents the Euclidean distance between the tracked target and other matching targets, and a1 and a2 represent preset weighting parameters.

[0080] In one embodiment, the matching values ​​of the tracked target and other unmatched targets are calculated respectively, the maximum value among the calculated matching values ​​of the tracked target is obtained, and it is determined whether the matching value is greater than a preset matching threshold. Here, if it is determined that the matching value is greater than the preset matching threshold, it is determined that the matching target corresponding to the matching value is successfully matched with the tracked target. If it is determined that the matching value is less than or equal to the preset matching threshold, it is determined that the tracking target is unsuccessfully matched, and the tracked target is used as another new target.

[0081] In summary, the target tracking device provided in the above embodiments acquires image information of the tracked target through a data acquisition module, performs a first-level matching of multiple targets based on a strong target tracking method through a first-level tracking module, performs a second-level matching of unmatched targets through a second-level tracking module based on a weak target tracking method, and achieves final matching and tracking through a third-level tracking module based on the parameter calculation method of the tracking module. This helps to improve the matching rate of multi-target tracking and matching, and enables more accurate target tracking.

[0082] Based on the same inventive concept as the foregoing embodiments, this invention provides a target tracking device, such as... Figure 3 As shown, the device includes: a processor 210 and a memory 211 storing a computer program; wherein, Figure 3 The processor 210 shown in the diagram does not refer to a single processor 210, but rather to its positional relationship relative to other devices. In practical applications, there can be one or more processors 210. Figure 3The memory 211 illustrated herein has the same meaning, that is, it is only used to indicate the positional relationship of memory 211 relative to other devices. In practical applications, there can be one or more memories 211. When the processor 210 runs the computer program, the target tracking method applied to the above-described device is implemented.

[0083] The device may also include at least one network interface 212. The various components of the device are coupled together via a bus system 213. It is understood that the bus system 213 is used to implement communication between these components. In addition to a data bus, the bus system 213 also includes a power bus, a control bus, and a status signal bus. However, for clarity, in... Figure 3 The general designated all buses as Bus System 213.

[0084] The memory 211 can be volatile or non-volatile, or both. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), ferromagnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disc, or compact disc read-only memory (CD-ROM); magnetic surface memory can be disk storage or magnetic tape storage. Volatile memory can be random access memory (RAM), used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Synchronous Static Random Access Memory (SSRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic Random Access Memory (SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM), SyncLink Dynamic Random Access Memory (SLDRAM), and Direct Rambus Random Access Memory (DRRAM).The memory 211 described in the embodiments of the present invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0085] The memory 211 in this embodiment of the invention is used to store various types of data to support the operation of the device. Examples of such data include: any computer programs used to operate on the device, such as operating systems and applications; contact data; phonebook data; messages; pictures; videos, etc. The operating system includes various system programs, such as the framework layer, core library layer, driver layer, etc., used to implement various basic services and handle hardware-based tasks. Applications can include various applications, such as media players, browsers, etc., used to implement various application services. Here, the program implementing the method of this embodiment of the invention can be included in the application.

[0086] Based on the same inventive concept as the foregoing embodiments, this embodiment also provides a computer-readable storage medium storing a computer program. The computer-readable storage medium can be a magnetic random access memory (FRAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM), etc.; it can also be various devices including one or any combination of the above-mentioned memories, such as mobile phones, computers, tablet devices, personal digital assistants, etc. When the computer program stored in the computer-readable storage medium is executed by a processor, it implements the target tracking method applied to the above-mentioned device. For the specific steps implemented when the computer program is executed by the processor, please refer to [link to relevant documentation]. Figure 1 The description of the illustrated embodiments will not be repeated here.

[0087] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0088] In this document, the terms “comprising,” “including,” or any other variations thereof are intended to cover non-exclusive inclusion, which includes not only the elements listed but also other elements not expressly listed.

[0089] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be defined by the scope of the claims.

Claims

1. A target tracking method, characterized in that, The method includes the following steps: Acquire image data and identify the tracking target from the image data; According to the preset target strong tracking method, the tracking target is strongly tracked and matched with other targets. If the strong tracking match of the tracking target is successful, the behavior trajectory of the tracking target is obtained. If the strong tracking match of the tracked target is unsuccessful, the tracked target is conditionally divided, and based on the result of the conditional division, the tracked target is weakly tracked and matched with other unmatched targets using a preset weak tracking method; if the weak tracking match of the tracked target is successful, the behavioral trajectory of the tracked target is obtained. If the weak tracking match of the tracked target is unsuccessful, the trajectory parameters of the tracked target are obtained, and the matching value of the tracked target with other unmatched targets is calculated based on the trajectory parameters. The trajectory of the tracked target is then supplemented with matching based on the matching value. The acquisition of image data and the determination of the tracking target from the image data include: Collect image data information, and perform target recognition and extraction on the image data information to obtain at least one candidate target; Based on the area size of all or some of the candidate targets, the targets are classified into multiple levels, and the target type of each level of candidate targets is identified. Based on the identification results of the target type, determine whether the candidate target for each level is a tracking target.

2. The method according to claim 1, characterized in that, The step of performing target recognition and extraction on the image data information to obtain at least one candidate target specifically includes: At least one initial target is obtained through the target identification and extraction process; Determine whether the percentage of each initial target in the total number of initial targets is greater than a preset percentage threshold; If so, then the initial target is determined to be a candidate target; If not, then the initial target is determined to be a candidate target.

3. The method according to claim 1, characterized in that, The step of determining whether a candidate target at each level is a tracking target based on the target type identification result also includes: Identify the type of the target at each level of the classification, determine whether the target is incorrectly identified based on the actual object of the target; obtain the detection error rate of the incorrect identification of the target type, and determine whether the detection error rate is greater than a preset error threshold; If the target detection error rate is determined to be greater than a preset error threshold, then the target with the highest error rate value is identified as the tracking target. If the target detection error rate is determined to be less than or equal to the preset error threshold, the type of the target is re-identified.

4. The method according to claim 1, characterized in that, If the strong tracking match of the tracked target fails, the tracked target is conditionally divided, and based on the result of the conditional division, the tracked target is weakly matched with other unmatched targets using a preset weak tracking method, including: If the tracked target fails to match other unmatched targets through strong target tracking, the area of ​​the tracked target is compared with a preset area benchmark value. If the area of ​​the tracked target exceeds the preset area benchmark value, the tracked target is divided into large targets, and the trajectory of the tracked target is tracked and matched based on the target feature comparison method. If the area of ​​the tracked target does not exceed the preset area benchmark value, the tracked target is divided into small targets, and the trajectory of the tracked target is tracked and matched based on the filter-based tracking method.

5. The method according to claim 1, characterized in that, If the weak tracking match of the tracked target fails, the trajectory parameters of the tracked target are obtained, and the matching value of the tracked target with other unmatched targets is calculated based on the trajectory parameters, including: If the tracked target fails to match other unmatched targets through weak tracking, feature extraction is performed on the tracked target and on the target corresponding to each trajectory to determine the similarity S of the tracked target; the Euclidean distance L between the center point of the tracked target and the center points of other unmatched targets in a frame of image data is calculated. And calculate the matching value of the tracked target based on the similarity S and the Euclidean distance L: P = a1*S + a2*L Wherein, P represents the matching value of the tracked target, S represents the similarity of the tracked target, L represents the Euclidean distance between the tracked target and other matching targets, and a1 and a2 represent preset weighting parameters.

6. The method according to claim 5, characterized in that, The step of supplementing the trajectory of the tracked target based on the matching value includes: Calculate the matching value of the tracked target with other unmatched targets respectively, obtain the maximum value among the calculated matching values ​​of the tracked target, and determine whether the matching value is greater than a preset matching threshold; If the matching value is determined to be greater than the preset matching threshold, then the matching target corresponding to the matching value is determined to be successfully matched with the tracking target; If the matching value is determined to be less than or equal to the preset matching threshold, then the tracking target is determined to be unsuccessful in matching, and the tracking target is used as another new target.

7. A target tracking device, characterized in that, The device includes: The data acquisition module is used to acquire image data and determine the tracking target from the image data; The first-level tracking module is used to perform strong tracking matching between the tracked target and other targets according to a preset strong tracking method. If the strong tracking matching of the tracked target is successful, the behavioral trajectory of the tracked target is obtained. The secondary tracking module is used to perform conditional classification on the tracking target if strong tracking matching of the tracking target fails, and to perform weak tracking matching of the tracking target with other unmatched targets according to the result of conditional classification using a preset weak tracking method; if the weak tracking matching of the tracking target is successful, the behavioral trajectory of the tracking target is obtained. The three-level tracking module is used to obtain the trajectory parameters of the tracking target if the weak tracking match of the tracking target fails, calculate the matching value of the tracking target with other unmatched targets based on the trajectory parameters, and perform supplementary matching of the trajectory of the tracking target based on the matching value; The data acquisition module is used to acquire image data information, perform target recognition and extraction on the image data information to obtain at least one candidate target; classify the candidate targets into levels according to the area size of all or some of the candidate targets, divide all or some of the candidate targets into multiple levels, and identify the target type of the candidate targets in each level; determine whether the candidate targets in each level are tracking targets based on the target type identification results.

8. An electronic device, characterized in that, include: A processor and a memory; wherein the memory stores a computer program for loading and executing the target tracking method as claimed in any one of claims 1 to 6 by the processor.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions for loading by a processor and executing the target tracking method as claimed in any one of claims 1 to 6.

Images