A multi-tracking device cooperative group target tracking method

Abstract

The application provides a multi-tracking device cooperative group target tracking method, which comprises the following steps: initial guiding, guiding the tracking device to an initial position to wait for a target to enter a field of view; target labeling, labeling the target according to target information; adaptive task allocation, allocating all targets in the field of view to the tracking device; guiding the tracking device to complete tracking of the target through a multi-tracking device cooperative working mode; and when a single task is completed, the tracking device participating in the current task is allocated to a new task. The application solves two problems in the group target tracking process, that is, a group target task planning problem and a device cooperative working mode design problem. The multi-device cooperative algorithm improves the working capacity of the system.

Description

Technical Field

[0001] This invention relates to the fields of collaborative tracking technology and distributed multi-agent collaborative technology, specifically to a group target tracking method involving multiple tracking devices. Background Technology

[0002] With the development of tracking technology, more and more devices are being used for target tracking, and tracking tasks are becoming increasingly complex. The task formats ranging from single-target to multi-target tracking have specific application scenarios in various fields. However, the increase in the number of targets leads to two problems:

[0003] First, the traditional working method of a single tracking device can no longer meet the needs;

[0004] Secondly, when there are multiple devices in the tracking system, how can the resources of all devices be integrated?

[0005] The collaborative working method is similar to the distributed working mode of sensor networks. All sensors transmit information throughout the network through information diffusion. Each sensor processes and fuses information locally by combining information from all its neighbors, thus determining its subsequent tasks. Distributed networks have three main advantages:

[0006] First, the concept of a central unit is removed in a distributed network. Tasks that were originally completed by the central unit are distributed to various local units, which reduces the computing resources of the central unit.

[0007] The robustness of the second network will increase, and the failure of a single node will not have a significant impact on the network, because other neighbors can take over and complete the subsequent work.

[0008] The third network has enhanced scalability. Distributed network operation mainly relies on communication between nodes. When a new node joins the network, it can know what work it needs to do by communicating with its neighbors, without the need for human intervention. All of the above advantages of distributed operation are also present in device collaborative tracking.

[0009] However, to introduce the concept of distributed systems into group target tracking tasks, two main issues need to be addressed.

[0010] First, there's the issue of task planning. With multiple devices tracking multiple targets, the challenge lies in assigning each target to the most suitable tracking device. Different targets have different characteristics, leading to variations in performance across different types of tracking devices. Therefore, assigning appropriate tracking targets based on the performance of the tracking device is a crucial aspect of task allocation.

[0011] Secondly, there is the issue of device coordination. Once the target is divided, devices assigned to the same target need to communicate and cooperate to ensure the successful completion of the task. In this patent, we designed a multi-device collaborative group target tracking algorithm, which can effectively solve the above two problems and improve the system's working capacity. Summary of the Invention

[0012] The purpose of this invention is to provide a multi-device collaborative group target tracking method for multi-target collaborative tracking. Through task planning, it realizes the organic integration of resources of multiple tracking devices and improves the system's working capacity.

[0013] The technical solution adopted in this invention is: a group target tracking method with multi-tracking device collaboration, comprising the following steps:

[0014] (1) Initial guidance: In the initial stage of the mission, each device guides itself to the initial position according to the guidance position and waits for the target to enter the field of view of the device detector;

[0015] (2) Target labeling: When multiple targets need to be detected, there are often multiple targets in the field of view. It is necessary to label the targets according to the target information and share the information of the target that has been labeled in the tracking device.

[0016] (3) Adaptive task allocation: Through the designed adaptive task allocation mechanism, all targets in the field of view can be effectively allocated to the tracking device;

[0017] (4) By using a multi-device collaborative working method, guide the equipment to complete the tracking of the target;

[0018] (5) When a single task is completed, the device that participated in the current task will be assigned a new task.

[0019] Furthermore, in step (1), the device aligns itself with the target direction based on the initial guidance information and waits for the target to enter the device's field of view.

[0020] Furthermore, in step (2), in the group target tracking task, in order to effectively combine multiple tracking devices organically, labeling the target is the basis for subsequent task allocation. In this patent, we designed a target labeling mechanism based on the target's location information and image information.

[0021] Furthermore, in step (3), the task allocation mechanism is based on the target labeling mechanism in the previous step. In this patent, we reasonably allocate tracking devices to the target based on information such as equipment performance and target status, thereby improving the equipment's ability to track the target.

[0022] Furthermore, in step (4), in the previous step, all targets were assigned to tracking devices. At this time, devices assigned to the same target will form a temporary group to complete the target tracking in a collaborative manner. The collaborative tracking method can increase the scalability of the entire system. When a new device joins the work, it can join the work simply by communicating with adjacent devices, without the need for manual task allocation. Secondly, the multi-device collaborative working method can effectively increase the robustness of the system.

[0023] Furthermore, in step (5), during group target tracking, different tracking devices will be assigned to different tracking targets. After the target assignment is completed, the devices forming temporary groups will coordinate to track the common task target through mutual communication. However, when the current tracking task ends, the tracking tasks of the other groups may still be in progress. At this time, the devices need to find new targets, join new groups, and complete a new round of tracking tasks. The collaborative working method can effectively improve the working capacity of the entire system and also greatly improve the adaptability of the system.

[0024] The advantages of this invention compared to the prior art are:

[0025] (1) The concept of collaboration is introduced into group target tracking, which effectively improves the system’s working capacity and scalability.

[0026] (2) Considering the different characteristics of the target and the tracking equipment, a task planning mechanism was designed.

[0027] (3) It improved the robustness of the original system and the adaptability of the equipment. Attached Figure Description

[0028] Figure 1 This diagram illustrates multi-device tracking. It assumes there are five tracking devices and two independent targets in a single task. The five devices are divided into two groups to track the targets collaboratively.

[0029] Figure 2 The algorithm flowchart represents the algorithm process in the form of a flowchart, which illustrates the steps of task planning in the algorithm, and corresponds in detail to algorithm step (3).

[0030] Figure 3 Through simulation Figure 1 The tracking accuracy of the tracking methods shown is as follows: Aim2 (the upper curve) represents the tracking accuracy when two devices are tracking, and Aim1 (the lower curve) represents the tracking accuracy when three devices are tracking. Detailed Implementation

[0031] Figure 1 This is a schematic diagram of multi-device tracking. (For example...) Figure 1 As shown, there are five tracking devices. In one mission, there are two independent targets that need to be tracked separately. The five devices form two groups to track the targets collaboratively.

[0032] The present invention proposes a multi-tracking device collaborative group target tracking method, the specific steps of which are as follows:

[0033] (1) Initial guidance;

[0034] All equipment is guided to the initial mission position based on the initial information, waiting for the target to enter the equipment's field of view.

[0035] (2) Target label;

[0036] When the target is within the field of view, there will be a miss distance. Using this miss distance data, the target's position in the horizontal coordinate system is calculated. Then, based on the equipment site information, the target's position in the geocentric coordinate system is calculated. An elevation communication flag is then set for the target whose position has been identified.

[0037] (3) Multi-objective task planning; such as Figure 2 The flowchart shown represents the method process in the form of a flowchart, which illustrates the steps of task planning in the method, specifically corresponding to step (3) of the method.

[0038] The devices communicate with each other. When a target is seen in the field of view of a device, it first calculates the target's position in the geocentric coordinate system. At the same time, by comparing with the communication between adjacent devices, it identifies whether the target is already being tracked by another device. There are several ways to process the task allocation based on the identification results:

[0039] 1) When it is found that the target is not being tracked by any equipment

[0040] If it is found that no device is tracking the current target, then step 2 is executed to tag the current target and notify the device to track the target;

[0041] 2) When it is detected that the target is being tracked by equipment

[0042] When a target is detected being tracked by the current device, the device needs to determine whether the target's tracking device is more suitable than itself to complete the tracking. To this end, we designed a scoring function for the current target's fitness, as follows:

[0043]

[0044] in, This indicates the current tracking score of the current target by the current device. It indicates target information, including target brightness, target area, etc.

[0045] If the current device's score is found to be higher than that of an existing tracking device, step 2 is executed to update the target's label information. When a device detects that the label information of the target it is tracking has been changed, it searches for the next target in its field of view. If the score is lower than that of an existing tracking device, it searches for the next target.

[0046] (4) Collaborative target tracking

[0047] In the process of cooperative target tracking, we use the Kalman algorithm as the basis for our research. The data update process involves the following steps:

[0048] 1) Algorithm parameter initialization, which involves... as well as Parameter initialization;

[0049]

[0050] in, This represents the state estimate of the system at the previous moment. This represents the covariance of the state estimate at the current moment. This indicates an intermediate estimate from the system, and this parameter will be updated in subsequent steps;

[0051] 2) Incremental update process: Multiple tracking devices share data through communication, and each tracking device updates its local parameters based on the acquired data;

[0052]

[0053] in, This refers to a tracking team composed of equipment with a work tracking target. Let represent the measurement noise covariance of detector l at time i. The measurement matrix of detector l at time i represents the conversion of state variables into observations. This represents the observation value of detector l at time i; through Find , The group of detectors that communicate with detector k is called the neighbor of detector k. -1 represents the transpose of a matrix, and -1 represents the inverse of a matrix.

[0054] 3) Data fusion: Each detector shares local parameters in a distributed network, while simultaneously acquiring local parameters from other sensors and performing data fusion.

[0055]

[0056] The data fusion coefficient represents the coefficient that satisfies:

[0057]

[0058] 4) Algorithm parameters , renew:

[0059]

[0060] in, This represents the state transition matrix at the current moment. This represents the state noise transition matrix at the current time. The covariance matrix is ​​the covariance of the state noise at the current moment.

[0061] The method of this invention introduces the concept of distributed collaboration into multi-target tracking, and designs a task planning method for multi-target tracking, which can effectively improve the stability and working capacity of the system. (See attached diagram) Figure 3 As shown, simulation was performed. Figure 1 The tracking accuracy of the tracking methods shown is as follows: Aim2 (the upper curve) represents the tracking accuracy when two devices are tracking, and Aim1 (the lower curve) represents the tracking accuracy when three devices are tracking.

[0062] 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 modifications or substitutions that can be conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of the present invention. Contents not described in detail in this specification are prior art known to those skilled in the art.

Claims

1. A method for group target tracking using multi-tracking devices in coordination, characterized in that, The method comprises the following steps: (1) Initial guidance: In the initial stage of the mission, each tracking device guides itself to the initial position according to the initial guidance information and waits for the target to enter the field of view of the tracking device detector; (2) Target labeling: When performing multi-target detection, there are multiple targets in the field of view. It is necessary to label the targets according to the target information and share the completed labeling information in the tracking device. (3) Adaptive task allocation: Through the adaptive task allocation mechanism, all targets in the field of view are allocated to the tracking device; (4) By using a multi-tracking-device collaborative working method, the tracking devices are guided to complete the tracking of the target; (5) When the current task is completed, the tracking equipment involved in the current task will be assigned a new task; Step (3) specifically involves: communication between devices. When a device sees a target in its field of view, it first calculates the target's position in the geocentric coordinate system. Simultaneously, by comparing with the communication between adjacent devices, it identifies whether the current target has already been tracked by any device. There are several processing methods for task allocation based on the identification results: 1) When it is found that the target is not being tracked by any equipment If it is found that no device is tracking the current target, then step (2) is executed to tag the current target and notify the device to track the target; 2) When it is detected that the target is being tracked by equipment When a target is detected being tracked by the current device, the device needs to determine whether the target's tracking device is more suitable than itself to complete the tracking of that target. The fitness score function for the current target is as follows: in, This indicates the current tracking score of the current target by the current device. This indicates target information, including: target brightness and target area information; If the current device's score is found to be higher than that of an existing tracking device, then step (2) is executed to update the label information of the target. When a device finds that the label information of the target it is tracking has been changed, it searches for the next target in the field of view. If the score is lower than that of the current tracking device, then it searches for the next target.

2. The method for group target tracking with multi-tracking device collaboration according to claim 1, characterized in that: In step (1), the tracking device aligns itself with the target direction based on the initial guidance information and waits for the target to enter the field of view of the tracking device.

3. The method for group target tracking with multi-tracking device collaboration according to claim 1, characterized in that: In step (2), a target labeling mechanism is designed based on the target's location information and image information.

4. The method for group target tracking with multi-tracking device collaboration according to claim 1, characterized in that: In step (3), a tracking device is assigned to the target based on the performance of the tracking device and the target status.

5. The method for group target tracking with multi-tracking device collaboration according to claim 1, characterized in that: In step (4), all targets are assigned to tracking devices. At this time, tracking devices assigned to the same target will form a temporary group to complete the tracking of the target in a collaborative manner.

6. A method for group target tracking with multi-tracking device collaboration according to claim 5, characterized in that: In step (5), during group target tracking, different tracking devices are assigned to their respective tracking targets. After the target assignment is completed, the tracking devices forming a temporary group work together to track the common task target by communicating with each other. When the current tracking task ends, the tracking tasks of the other groups are still in progress. At this time, the tracking devices need to find new targets, join the new group, and complete a new round of tracking tasks.

Images