A multi-target tracking method, system, electronic device and storage medium

Abstract

The present disclosure relates to the technical field of computer vision, and discloses a multi-target tracking method, a system, an electronic device and a storage medium, wherein the multi-target tracking method comprises: acquiring a current image at a current moment and a previous image at a previous moment; the current image comprises a plurality of current targets, and the previous image comprises a plurality of previous targets; for each current target, determining a difference value of the current target and each previous target; and determining a tracking target corresponding to each current target from the previous targets according to the difference value of the current target and each previous target. Through the present disclosure, the tracking accuracy of multiple current targets can be improved, and the occurrence of confusion and omission and the like is reduced, thereby solving the problem of poor target tracking effect in the prior art.

Description

TECHNICAL FIELD

[0001] The present disclosure belongs to the technical field of computer vision, and particularly relates to a multi-target tracking method, system, electronic device and storage medium. BACKGROUND

[0002] The multi-target tracking method has wide application prospects in automatic driving, intelligent transportation systems, security and other fields. The method can accurately detect and track multiple targets in the surrounding environment.

[0003] However, in complex and variable practical application scenarios, traditional multi-target tracking technology has some challenges and limitations. For example, in high-density target scenarios, traditional multi-target tracking algorithms face problems such as target confusion, misassociation, and omission, and the tracking accuracy decreases; in cases where the target speed changes greatly or the targets cross, the traditional algorithm cannot accurately determine the motion state of the target, resulting in unsatisfactory tracking results; in addition, the traditional algorithm may also be affected by factors such as complex environmental interference and target occlusion, resulting in poor real-time tracking.

[0004] The existing related technologies still cannot well solve the problem of poor target tracking effect. SUMMARY

[0005] To solve the above problems, the present disclosure provides a multi-target tracking method, system, electronic device and storage medium, which solves the problem of poor target tracking effect in the existing related technologies.

[0006] The present disclosure provides a multi-target tracking method, which comprises:

[0007] obtaining a current image at a current time and a previous image at a previous time; the current image comprises a plurality of current targets, and the previous image comprises a plurality of previous targets;

[0008] For each current target, determining a difference value between the current target and each previous target;

[0009] According to the difference value between the current target and each previous target, determining a tracking target corresponding to each current target from the previous targets.

[0010] In some embodiments, determining the difference value between the current target and each previous target comprises:

[0011] determining an index difference value between the current target and the previous target; the index difference value comprises a horizontal coordinate difference value, a vertical coordinate difference value, and a speed difference value;

[0012] The difference between the current target and the previous target is brought into a preset evaluation function to determine the difference value between the current target and the previous target.

[0013] Preferably, the expression of the evaluation function is as follows:

[0014] Fun(Δx,Δy,Δv)=αΔx 2 +βΔy 2 +γΔv 2 ;

[0015] Wherein, α, β, γ are weight coefficients, α+β+γ=1, Δx is the horizontal coordinate difference, Δy is the vertical coordinate difference, and Δv is the speed difference.

[0016] Preferably, determining the tracking target corresponding to each current target from the previous targets comprises:

[0017] For each current target, the previous target with the minimum difference value with the current target is selected as the tracking target corresponding to the current target; the tracking target corresponding to each current target is not the same.

[0018] In some embodiments, selecting the previous target with the minimum difference value with the current target as the tracking target corresponding to the current target comprises:

[0019] According to the difference value between the current target and the previous target, an evaluation function matrix is constructed; the values in the same row in the evaluation function matrix are the difference values between the same current target and all previous targets, and the values in the same column in the evaluation function matrix are the difference values between the same previous target and all current targets;

[0020] Iteratively selecting the minimum value in the evaluation function matrix and updating the evaluation function matrix to determine a target set;

[0021] According to the target set, the tracking target corresponding to each current target is determined.

[0022] Preferably, selecting the minimum value in the evaluation function matrix and updating the evaluation function matrix comprises:

[0023] Selecting the minimum value in the evaluation function matrix;

[0024] Deleting the column where the minimum value is located to update the evaluation function matrix.

[0025] Preferably, before determining the tracking target corresponding to each current target according to the target set, it comprises:

[0026] The difference value of each item in the target set is compared with a preset threshold value, and if the difference value is greater than the preset threshold value, the difference value is removed from the target set.

[0027] The present disclosure provides a multi-target tracking system, comprising:

[0028] An acquisition module is configured to acquire a current image at a current time and a previous image at a previous time; the current image comprises a plurality of current targets, and the previous image comprises a plurality of previous targets;

[0029] A processing module is configured to determine, for each current target, a difference value between the current target and each previous target.

[0030] A determination module is configured to determine, according to the difference value between each current target and each previous target, a tracking target corresponding to each current target from the previous targets.

[0031] The present disclosure provides an electronic device, comprising:

[0032] A processor and a memory;

[0033] The processor invokes a computer program stored in the memory to execute the multi-target tracking method of the first aspect.

[0034] The present disclosure provides a computer-readable storage medium having a computer program stored therein, which, when executed by a processor, enables the processor to execute the multi-target tracking method of the first aspect.

[0035] The present disclosure has the following beneficial effects:

[0036] (1) The present disclosure provides a multi-target tracking method, which determines the situation or state of each current target in the previous image more comprehensively and accurately by determining the difference value of each current target in the current image, thereby improving the tracking accuracy of multiple current targets and reducing the occurrence of confusion and omission, and solving the problem of poor target tracking effect in the prior art.

[0037] (2) In the present disclosure, after any previous target is selected as the tracking target of a certain current target, the previous target cannot be selected again by other current targets, that is, the tracking target corresponding to each current target is different, and different current targets cannot be tracked by the same previous target, thereby preventing poor target tracking or confusion.

[0038] Other features and advantages of the present disclosure will be set forth in the description that follows, and in part will be apparent from the description, or can be learned by practice of the present disclosure. The purposes and other advantages of the present disclosure will be realized and attained by the structure particularly pointed out in the description and appended claims. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present disclosure, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0040] Figure 1 is a flow chart of the multi-target tracking method in the present disclosure;

[0041] Figure 2 is a schematic diagram of matching the current target and the previous target in the present disclosure;

[0042] Figure 3 is a structural block diagram of the multi-target tracking system in the present disclosure;

[0043] Figure 4 is a structural block diagram of the electronic device in the present disclosure. DETAILED DESCRIPTION

[0044] Example implementations will now be described more fully with reference to the accompanying drawings. Example implementations may, however, be implemented in many different forms and should not be construed as limited to the implementations set forth herein; rather, these implementations are provided so that this disclosure will be thorough and complete, and will fully convey the scope of example implementations to those skilled in the art. The described features, structures, or characteristics can be combined in one or more implementations. In the following description, numerous specific details are provided to give a thorough understanding of implementations of the present disclosure. One skilled in the relevant art will recognize, however, that the implementations of the present disclosure can be practiced without one or more of the specific details, or with other methods, components, materials, and so forth. In other instances, well-known structures have not been described in detail to avoid obscuring the aspects of the present disclosure.

[0045] Furthermore, the accompanying drawings are included to provide a further understanding of the present disclosure and are incorporated in and constitute a part of this specification. The drawings are not intended to be restrictive in any way. Like reference numerals in the various drawings are intended to represent the same or similar elements.

[0046] The flowcharts shown in the drawings are merely illustrative and do not necessarily include all the steps. For example, some steps can be further divided, and some steps can be combined or partially combined, so the actual execution order can be changed according to the actual situation.

[0047] The terms "first", "second", and the like in the specification and claims of the present application and the above drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.

[0048] In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or sub-modules does not necessarily have to be limited to those steps or sub-modules clearly listed, but can include other steps or sub-modules that are not clearly listed or inherent to these processes, methods, products or devices.

[0049] The present disclosure provides a multi-target tracking method, Figure 1 is a flowchart of the multi-target tracking method in the present disclosure, as Figure 1 shown, the method comprises the following steps:

[0050] Step S101, obtaining a current image at a current time and a previous image at a previous time; the current image includes a plurality of current targets, and the previous image includes a plurality of previous targets.

[0051] Step S102, for each current target, determining a difference value of the current target and each previous target.

[0052] Step S103, determining a tracking target corresponding to each current target from the previous targets according to the difference value of the current target and each previous target.

[0053] For example, set the current time as time t, then the last time is time t-1. Obtain the current image and the previous image, wherein the current image contains N current targets, and the previous image contains M previous targets. For each current target, match the current target with each previous target to determine the difference value of the current target and each previous target, as shown in the following formula. Figure 2 Figure 2 is a schematic diagram of matching the current target with the previous target in the present disclosure. The difference value can represent the difference between the current target and the corresponding previous target in different dimensions, and the smaller the difference value, the more likely the current target is tracked from the previous target. After determining the difference value for each current target, the tracking target corresponding to the current target is selected from the previous targets according to the difference value. Generally, the previous target with the smallest difference value is selected as the tracking target.

[0054] In the present method, the difference value is determined for each current target in the current image, which can more comprehensively and accurately determine the situation or state of the current target at the last time, thereby improving the tracking accuracy of multiple current targets and reducing the occurrence of confusion and omission, and solving the problem of poor target tracking effect in the related art.

[0055] The difference value is mainly used to reflect the degree of change of the position or state of the current target relative to the previous target, and the specific determination method or determination basis can be selected according to the dimension focused in the actual application scenario. In some embodiments, determining the difference value of the current target and each previous target includes: determining an index difference value of the current target and the previous target; the index difference value includes a horizontal coordinate difference value, a vertical coordinate difference value, and a speed difference value; and bringing the index difference value of the current target and the previous target into a pre-set evaluation function to determine the difference value of the current target and the previous target. In this embodiment, the horizontal coordinate difference value, the vertical coordinate difference value, and the speed difference value can help determine the position change process and the final state of the corresponding current target relative to the previous target.

[0056] Specifically, the expression of the evaluation function is as follows:

[0057] Fun(Δx,Δy,Δv)=αΔx 2 +βΔy 2 +γΔv 2 ;

[0058] wherein α, β, and γ are weight coefficients, α+β+γ=1, Δx is the horizontal coordinate difference value, Δy is the vertical coordinate difference value, and Δv is the speed difference value. The values of the three weight coefficients α, β, and γ can be adjusted according to the actual application scenario, and in this embodiment, α is 0.2, β is 0.5, and γ is 0.3. ​

[0059] In some embodiments, determining the tracking target corresponding to each current target from the previous targets comprises: for each current target, selecting the previous target with the minimum difference value from the current target as the tracking target corresponding to the current target; and the tracking target corresponding to each current target is different.

[0060] In the embodiment, the minimum difference value indicates that the current target is most likely the same target as the previous target or is adjusted from the previous target, and thus selecting the previous target as the tracking target corresponding to the current target can improve the accuracy of target tracking. It is worth noting that the tracking target corresponding to each current target is different, i.e., different current targets cannot be tracked by the same previous target, which prevents poor target tracking or confusion.

[0061] Specifically, selecting the previous target with the minimum difference value from the current target as the tracking target corresponding to the current target comprises: constructing an evaluation function matrix according to the difference values between the current targets and the previous targets; the values in the same row in the evaluation function matrix are the difference values between the same current target and all previous targets, and the values in the same column in the evaluation function matrix are the difference values between the same previous target and all current targets; iteratively selecting the minimum value in the evaluation function matrix and updating the evaluation function matrix to determine a target set; and determining the tracking target corresponding to each current target according to the target set.

[0062] For example, for N current targets at time t and M previous targets at time t-1, after the difference values are determined, an evaluation function matrix of n rows and m columns is constructed, F[n][m] The values in the same row in the evaluation function matrix are the difference values between the same current target and all previous targets, and the values in the same column in the evaluation function matrix are the difference values between the same previous target and all current targets. According to the evaluation function matrix, the difference values between any current target and all previous targets can be more intuitively represented, which is helpful for determining the tracking target of the current target.

[0063] To avoid the situation that multiple current targets select the same previous target as the tracking target, more specifically, selecting the minimum value in the evaluation function matrix and updating the evaluation function matrix comprises: selecting the minimum value in the evaluation function matrix; and deleting the column where the minimum value is located to update the evaluation function matrix.

[0064] For example, for the above evaluation function matrix, the minimum value in the matrix is first selected and denoted as f ij, indicating that the i-th current target at time t is tracked from the j-th previous target at time t-1, and the current target is the current target in the current evaluation function matrix that is most easily determined to track the target. Since the previous target has been selected, in order to prevent other current targets from selecting the previous target again in the subsequent process, the column where the minimum value is located is deleted from the evaluation function matrix, that is, the difference value between the previous target and all current targets is deleted, realizing the update of the evaluation function matrix. Then, the minimum value in the updated evaluation function matrix is selected again, and so on, until all the previous targets corresponding to the current targets are selected.

[0065] According to the minimum value selected each time in the above process, a target set is constructed, and the previous target corresponding to each minimum value in the set is recorded as the tracking target of the corresponding current target, and the tracking target of each current target is different.

[0066] In the actual tracking process, from time t-1 to time t, there may be a possibility that targets are added or targets are not in the monitoring range. If n>m, it indicates that there are new targets at the current time, and if n

[0067] In the embodiment, if the difference value in the target set is greater than the preset threshold, it indicates that the difference between the current target and the previous target in the difference value is too large, so that the possibility of tracking or being tracked between the two is exceeded. It is explained that the current target corresponding to the difference value is a new target, or the previous target corresponding to the difference value is not in the detection range at the current time. Therefore, the difference value greater than the preset threshold is removed from the target set, and the current target corresponding to the difference value is not tracked. The preset threshold can be set according to actual needs, and in the embodiment, the preset threshold is set to 0.5.

[0068] In summary, in the method, the difference value of each current target in the current image is determined, which can more comprehensively and accurately determine the situation or state of the current target at the previous time, thereby improving the tracking accuracy of multiple current targets, reducing the occurrence of confusion and omission, and solving the problem of poor target tracking effect in the existing related technology. And in the method, the tracking target of each current target is finally determined to be different, which can avoid the occurrence of tracking confusion.

[0069] The present disclosure also provides a multi-target tracking system, Figure 3 is a structural block diagram of the multi-target tracking system in the present disclosure, as Figure 3 shown, the system comprises:

[0070] The acquisition module 301 is configured to acquire a current image at a current time and a previous image at a previous time; the current image comprises a plurality of current targets, and the previous image comprises a plurality of previous targets.

[0071] The processing module 302 is configured to determine, for each current target, a difference value of the current target and each previous target.

[0072] The determination module 303 is configured to determine, according to the difference value of each current target and each previous target, a tracking target corresponding to each current target from the previous targets.

[0073] In the present system, the difference value is determined for each current target in the current image, which can more comprehensively and accurately determine the situation or state of the current target at the previous time, thereby improving the tracking accuracy of multiple current targets and reducing the occurrence of confusion and omission, and solving the problem of poor target tracking effect in the prior art.

[0074] As shown in Figure 4 , Figure 4 is a structural block diagram of the electronic device in the present disclosure, and the present disclosure also provides an electronic device corresponding to the multi-target tracking method provided above. Since the embodiments of the device are similar to the method embodiments described above, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiments above. The device described below is only illustrative. The device can include a processor 1, a memory 2, a communication bus (i.e. the device bus) and a lookup engine. The processor 1 and the memory 2 can communicate with each other through the communication bus and communicate with the outside through the communication interface. The processor 1 can call the logical instructions in the memory 2 to execute the multi-target tracking method described above.

[0075] In addition, the logic instructions in the memory 2 described above can be implemented in the form of a software function unit and sold or used as an independent product, and can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present disclosure essentially or the part that contributes to the prior art or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for making a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods of various embodiments of the present disclosure. The aforementioned storage medium includes: a storage chip, a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0076] In another aspect, the embodiments of the present disclosure also provide a processor readable storage medium, and the processor readable storage medium stores a computer program 3. When the computer program 3 is executed by the processor 1, the multi-target tracking method provided by the above-mentioned embodiments is implemented.

[0077] The processor readable storage medium can be any available medium or data storage device that the processor 1 can access, including but not limited to a magnetic storage (such as a floppy disk, a hard disk, a magnetic tape, a magneto-optical disk (MO), etc.), an optical storage (such as a CD, a DVD, a BD, a HVD, etc.), and a semiconductor memory (such as a ROM, an EPROM, an EEPROM, a non-volatile memory (NAND FLASH), a solid state disk (SSD)), etc.

[0078] Those skilled in the art should understand that although the present disclosure is described in detail with reference to the foregoing embodiments, the technical solutions described in the foregoing embodiments can be modified, or some of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A multi-target tracking method, characterized in that, The method includes: Acquire the current image at the current moment and the preceding image at the previous moment; the current image includes several current targets, and the preceding image includes several preceding targets; For each current target, determine the difference between the current target and each preceding target, including: determining the index difference between the current target and the preceding target; the index difference includes the difference in the horizontal axis, the difference in the vertical axis, and the difference in velocity; substituting the index difference between the current target and the preceding target into a pre-set evaluation function to determine the difference between the current target and the preceding target; the expression of the evaluation function is as follows: ;in, , , All are weighting coefficients. , The difference is the x-axis value. The difference in the vertical axis. This is the speed difference; Based on the difference between the current target and each of the preceding targets, the tracking target corresponding to each current target is determined from the preceding targets.

2. The multi-target tracking method according to claim 1, characterized in that, Determining the tracking target corresponding to each current target from the preceding targets includes: For each current target, the preceding target with the smallest difference value from the current target is selected as the tracking target corresponding to the current target; the tracking targets corresponding to each current target are different.

3. The multi-target tracking method according to claim 2, characterized in that, Selecting the preceding target with the smallest difference value from the current target as the tracking target corresponding to the current target includes: An evaluation function matrix is ​​constructed based on the difference between the current target and the preceding target; the values ​​in the same row of the evaluation function matrix are the difference between the same current target and all preceding targets, and the values ​​in the same column of the evaluation function matrix are the difference between the same preceding target and all current targets. Iteratively select the minimum value in the evaluation function matrix and update the evaluation function matrix to determine the target set; The tracking target corresponding to each current target is determined based on the target set.

4. The multi-target tracking method according to claim 3, characterized in that, Selecting the minimum value in the evaluation function matrix and updating the evaluation function matrix includes: Select the minimum value in the evaluation function matrix; Delete the column containing the minimum value and update the evaluation function matrix.

5. The multi-target tracking method according to claim 3, characterized in that, Before determining the tracking target corresponding to each current target based on the target set, the process includes: Traverse the target set and compare the difference value of each item in the target set with a preset threshold. If the difference value is greater than the preset threshold, remove the difference value from the target set.

6. A multi-target tracking system, characterized in that, include: The acquisition module is used to acquire the current image at the current moment and the preceding image at the previous moment; the current image includes several current targets, and the preceding image includes several preceding targets; The processing module, for each current target, determines the difference between the current target and each preceding target, including: determining the index difference between the current target and the preceding target; the index difference includes the difference in the horizontal axis, the difference in the vertical axis, and the difference in velocity; and inputting the index difference between the current target and the preceding target into a pre-set evaluation function to determine the difference between the current target and the preceding target; the expression of the evaluation function is as follows: ;in, , , All are weighting coefficients. , The difference is the x-axis value. The difference in the vertical axis. This is the speed difference; The determination module is used to determine the tracking target corresponding to each current target from the preceding targets based on the difference value between the current target and each preceding target.

7. An electronic device, characterized in that, include: Processor and memory; The processor invokes the computer program stored in the memory to execute the multi-target tracking method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, enables the processor to perform the multi-target tracking method according to any one of claims 1 to 5.

Images