Regional target detecting and tracking method and system

Abstract

The invention provides a region target detection tracking method and system. The method comprises the following steps: acquiring an echo characteristic change state of a target region through a three-frequency microwave radar array; determining a change area through the change state; obtaining a point cloud picture of the change area through the three-dimensional laser radar array; splicing point cloud coordinates, adjacent to the change area, in the initial point cloud atlas with the point cloud coordinates of the change area to generate a supplementary change point cloud atlas; extracting feature data of the change target according to the supplementary change point cloud atlas; and if any item in the feature data of the change target is different from the initial target, newly adding the change target into a target list, and controlling the three-dimensional laser radar array and the camera array to track the change target so as to obtain the position and the image of the change target. According to the method, feature data of a changing target is extracted by supplementing a changing point cloud picture, and the changing target is accurately recognized by combining an initial target, so that the problem that the target in a monitoring area cannot be accurately detected and tracked is solved.

Description

Technical Field

[0001] The present application relates to the field of regional security, and in particular to a method and system for regional target detection and tracking. Background Art

[0002] In fields such as security, it is crucial to accurately and efficiently obtain characteristic information such as the size, position, and movement mode of the target in the monitoring area, which is the basis for further judging the target type.

[0003] Manual inspection, manual video monitoring inspection, image recognition monitoring inspection and vibration fiber optic monitoring inspection can be used. Among them, manual inspection uses patrol personnel to walk around, manual video detection inspection completes the inspection by patrol personnel monitoring the camera screen, image recognition monitoring inspection conducts inspection through image recognition, and vibration fiber optic monitoring inspection monitors the area through the installation of fences using vibration fiber optics.

[0004] However, manual inspections rely on the labor intensity of inspectors, have limited coverage, and are greatly affected by subjective factors. The camera images of manual video monitoring inspections are not intuitive, and the visual acuity of the human eye is limited. Image recognition monitoring inspections have limitations and are prone to misjudgment. Vibration fiber optic monitoring inspections require fences in the monitoring area and cannot know the target information in the area. Therefore, manual inspections, manual video monitoring inspections, image recognition monitoring inspections, and vibration fiber optic monitoring inspections cannot accurately detect and track targets in the monitoring area. Summary of the invention

[0005] The present application provides a method and system for regional target detection and tracking to solve the problem of being unable to accurately detect and track targets within a monitoring area.

[0006] In a first aspect, the present application provides a method for detecting and tracking regional targets, comprising: obtaining the echo characteristic change state of the target area through a three-frequency microwave radar array; if any item in the characteristic data of the changed target is different from the initial target, the changed target is added to the target list, and the three-dimensional laser radar array is controlled to determine the changed area through the change state, and the changed area includes multiple changed targets; obtaining a point cloud map of the changed area through a three-dimensional laser radar array, and the point cloud map includes multiple point cloud coordinates; splicing the point cloud coordinates adjacent to the changed area in the initial point cloud map with the point cloud coordinates of the changed area to generate a supplementary changed point cloud map, and the initial point cloud map includes a target list, and the target list includes multiple initial targets; extracting the characteristic data of the changed target according to the supplementary changed point cloud map, and the characteristic data includes target size, target position and target outline; the camera array tracks the changed target to obtain the position and image of the changed target.

[0007] Optionally, the method also includes: scanning the target point cloud coordinates of the target area by the three-dimensional laser radar, the three-dimensional laser radar array includes multiple three-dimensional laser radars; splicing the target point cloud coordinates into the same scene to generate an initial point cloud map of the target area; identifying the targets of the initial point cloud map to form a target list, the target list includes the feature data of each target; and acquiring the echo characteristics of the target area by a three-frequency microwave radar array.

[0008] Optionally, the method further includes: performing environmental filtering on the echo feature to obtain a filtered echo feature; and comparing the filtered echo feature with an initial echo feature to determine a change state of the filtered echo feature.

[0009] Optionally, the method further includes: if all items in the feature data of the changed target are the same as those of the initial target, setting the three-dimensional laser radar and camera in the changed area to an idle state.

[0010] Optionally, the method also includes: obtaining the coverage area of ​​the three-dimensional laser radar, the coverage area is divided into at least four categories, and the weight of each category of the coverage area is different; calculating the scheduling score of the three-dimensional laser radar, the scheduling score is the product of the weight of the coverage area and the area of ​​the coverage area; setting the scheduling priority of the three-dimensional laser radar according to the scheduling score, and starting the three-dimensional laser radar according to the scheduling priority.

[0011] Optionally, the target area is a polygonal area; the method also includes: if the target area is not a polygonal area, dividing the target area into multiple polygonal sub-areas according to the area of ​​the target area and the area of ​​the polygonal area, and setting the tracking range of the three-dimensional laser radar and the camera according to the sub-areas.

[0012] Optionally, the method also includes: if the change state is a change, determining a change area; determining the change area monitored by the three-frequency microwave radar unit according to the ID of the three-frequency microwave radar unit that outputs the change state, and the three-frequency microwave radar array includes multiple three-frequency microwave radar units.

[0013] Optionally, the method also includes: recording the initial position of the changing target; acquiring a multi-frame point cloud map of the target area through the three-dimensional lidar array; traversing the moving position of the changing target in the multi-frame point cloud map; calculating the moving distance according to the moving position and the initial position, and calculating the movement time according to the number of frames of the multi-frame point cloud map; calculating the movement speed of the changing target according to the moving distance and the movement time.

[0014] In the second aspect, the present application provides a regional target detection and tracking system, including: a three-frequency microwave radar array, a data processing module and a three-dimensional laser radar array, wherein the three-frequency microwave radar array is used to obtain the echo characteristic change state of the target area; the data processing module is used to determine the changed area through the change state; the point cloud coordinates adjacent to the changed area in the initial point cloud map are spliced ​​with the point cloud coordinates of the changed area to generate a supplementary change point cloud map; the feature data of the changed target is extracted according to the supplementary change point cloud map; if any item of the feature data of the changed target is different from the initial target, the changed target is added to the target list; the three-dimensional laser radar array and the camera array are controlled to track the changed target to obtain the position and image of the changed target; the three-dimensional laser radar array is used to obtain the point cloud map of the changed area, and is also used to track the changed target to obtain the position of the changed target; the camera array is used to track the changed target to obtain the image of the changed target.

[0015] Optionally, it includes: the three-frequency microwave radar array includes multiple three-frequency microwave radar units, and the three-frequency microwave radar units are arranged at the end points or inside the target area; the three-dimensional laser radar array includes multiple three-dimensional laser radars, and the three-dimensional laser radar includes at least one solid-state radar and at least one high-speed pan-tilt head, or includes at least one multi-line radar and at least one high-speed pan-tilt head, and the high-speed pan-tilt head is a two-dimensional radar; the camera array includes multiple cameras.

[0016] It can be seen from the above technical scheme that the present application provides a method and system for detecting and tracking regional targets, which obtains the echo characteristic change state of the target area through a three-frequency microwave radar array; and determines the change area through the change state; then obtains the point cloud map of the change area through a three-dimensional laser radar array; splices the point cloud coordinates adjacent to the change area in the initial point cloud map with the point cloud coordinates of the change area to generate a supplementary change point cloud map; then extracts the feature data of the change target according to the supplementary change point cloud map, and the feature data includes target size, target position and target outline; if any item of the feature data of the change target is different from the initial target, the change target is added to the target list, and the three-dimensional laser radar array and the camera array are controlled to track the change target to obtain the position and image of the change target. The method extracts the feature data of the change target through the supplementary change point cloud map, and accurately identifies the change target in combination with the initial target, so as to solve the problem of being unable to accurately detect and track the target in the monitoring area, and the detection and tracking problem of the target moving when multiple targets appear in the area and the target is in a blind spot. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solution of the present application, the drawings required for use in the embodiments are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0018] Figure 1 A schematic diagram of a method for detecting and tracking regional targets;

[0019] Figure 2 This is a schematic diagram of the target monitoring area;

[0020] Figure 3 It is a schematic diagram of target movement in the target monitoring area;

[0021] Figure 4 It is a schematic diagram of multiple target monitoring areas;

[0022] Figure 5 This is a schematic diagram of the target monitoring area segmentation;

[0023] Figure 6 A schematic diagram of the segmentation of the polygonal target monitoring area for the interactive traffic direction;

[0024] Figure 7 This is a schematic diagram of a three-frequency microwave radar array;

[0025] Figure 8 This is a schematic diagram of the installation location of the three-frequency microwave radar array;

[0026] Fig. 9 A schematic diagram of another installation position of a three-frequency microwave radar array;

[0027] Fig.10 This is a schematic diagram of the first composition of a three-dimensional laser radar;

[0028] Fig.11 This is a schematic diagram of the second composition of a three-dimensional laser radar;

[0029] Fig.12 This is a schematic diagram of the third composition of a three-dimensional laser radar;

[0030] Fig.13 This is a schematic diagram of the installation location of the three-dimensional laser radar;

[0031] Fig.14 It is a schematic diagram of the rotation direction of the three-dimensional laser radar;

[0032] Fig.15 A schematic diagram of the camera array installation position;

[0033] Fig.16 Schematic diagram of scheduling priorities for 3D lidar. DETAILED DESCRIPTION

[0034] To facilitate the technical solution of the application, some concepts involved in the application are first explained below.

[0035] The microwave radar can use a smaller antenna to obtain a very narrow beam width to obtain more information about the properties of the measured target. The three-frequency microwave radar in this embodiment can be a microwave radar with three different frequencies.

[0036] 3D LiDAR uses lasers to measure distance. A transmitter sends out pulses of light that reflect off surfaces and return to a receiver. The receiver records the time it takes for each light pulse to return, which is used to calculate the distance each beam travels. These times and distances produce a set of data points called a point cloud, which can be converted into a 3D image depicting surfaces, terrain, structures, and objects.

[0037] The following embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following embodiments do not represent all implementations consistent with the present application. They are only examples of systems and methods consistent with some aspects of the present application as detailed in the claims.

[0038] See also Figure 1 In order to solve the problem that manual inspection, manual video monitoring inspection, image recognition monitoring inspection and vibration optical fiber monitoring inspection cannot accurately detect and track targets in the monitoring area, some embodiments of the present application provide a method for regional target detection and tracking, including:

[0039] S100: Acquire the echo characteristic change state of the target area through the three-frequency microwave radar array.

[0040] The three-frequency microwave radar array includes a plurality of three-frequency microwave radar units. In some embodiments, the three-frequency microwave radar units operate separately in the monitoring area. Figure 2 If the monitoring area is a rectangle, refer to Figure 8 For example: a three-frequency microwave radar array such as Figure 8 The three-frequency microwave radar units in the three-frequency microwave radar array move from the surroundings of the monitoring area to the monitoring center, and record the echo characteristics of the corresponding irradiated area.

[0041] In some embodiments, the three-frequency microwave radar array can also be Fig. 9 The three-frequency microwave radar unit in the three-frequency microwave radar array moves from the inside of the monitoring area to the surrounding area of ​​the monitoring area, and records the echo characteristics of the corresponding irradiated area.

[0042] To reduce interference from weather factors, in some embodiments, when executing the step of acquiring the echo characteristic change state of the target area through a three-frequency microwave radar array, environmental filtering can also be performed on the echo characteristics to obtain filtered echo characteristics; the filtered echo characteristics are compared with the initial echo characteristics to determine the change state of the filtered echo characteristics, wherein the environmental filtering parameters are an empirical matrix.

[0043] S200: Determine a change region according to the change state.

[0044] The change area includes multiple change targets. The change state can be a change or no change. It is understandable that if the change state is no change, there is no need to execute the next step; if the change state is a change, the filtered echo feature can be compared with the initial feature. When a significant change occurs, the result of the change can be output to the data processing module.

[0045] The process of obtaining the initial features is as follows: first, after the three-frequency microwave radar array, three-dimensional laser radar array, camera array and data processing module are connected to the same local area network, the three-frequency microwave radar array, three-dimensional laser radar array, camera array and data processing module are initialized. During initialization, it is necessary to ensure that there are no moving targets in the target monitoring area. The specific process of initialization is as follows: each three-frequency microwave radar of the three-frequency microwave radar array works in turn, records the echo characteristics of the irradiated area as the initial features; the three-dimensional laser radar of the three-dimensional laser radar array covers and scans the monitoring area, so that each position in the monitoring area is scanned by at least one three-dimensional laser radar.

[0046] After initialization, the data processing module stitches the scanning results of the 3D laser radar into the same scene to generate an initial point cloud map; the data processing module controls the camera array to patrol the monitoring area and collect images of the monitoring area. It is understandable that during the initialization process, there may be moving targets in the target monitoring area. When there are moving targets, the initial point cloud map is analyzed to form a global target list.

[0047] It should be understood that the initial target is not only the target that exists in the initialization. When the changed area covers the initial point cloud map, the target that exists in the initial point cloud map after the default coverage is the initial target relative to the changed target.

[0048] In some embodiments, the area target detection and tracking method also includes: if the change state is changed, determining the change area; and then determining the change area monitored by the three-frequency microwave radar unit based on the three-frequency microwave radar unit ID that outputs the change state.

[0049] In the above embodiment, the changed area can be found according to the ID of the three-frequency microwave radar unit that outputs the changed state. After the changed area is found, the three-dimensional laser radar array is controlled by the data processing module.

[0050] S300: Acquire a point cloud image of the changed area through a three-dimensional laser radar array.

[0051] The point cloud image includes multiple point cloud coordinates; for example, if there are two targets in the change area, the point cloud image includes the point cloud coordinates of the two targets, and the point cloud coordinates of the first target can be (x 1 ,y 1 ,z 1 ), the point cloud coordinates of the second target can be (x 2 ,y 2 ,z 2 ).

[0052] S400: splicing the point cloud coordinates adjacent to the changed area in the initial point cloud map with the point cloud coordinates of the changed area to generate a supplemented changed point cloud map.

[0053] Among them, the initial point cloud includes a target list, and the target list includes multiple initial targets; for example: the monitoring area is divided into six target areas, namely area 1, area 2, area 3, area 4, area 5 and area 6. If the changed area is area 2, because the monitoring area is a polygonal area, the adjacent areas of area 2 can be area 1, area 3 and area 5, then the area 1, area 3 and area 5 in the initial point cloud are spliced ​​with the point cloud coordinates of the changed area, that is, the point cloud coordinates of area 2 in the initial point cloud are replaced with the point cloud coordinates of the changed area. For example: there is no point cloud coordinate in area 2 in the initial point cloud, and there are two targets in the changed area, and the point cloud coordinates are (x 1 ,y 1 ,z 1 ) and (x 2 ,y 2 ,z 2 ), therefore, there are two targets in region 2 in the supplementary change point cloud map, and the target list in the supplementary change point cloud map includes target 1 and target 2. The point cloud coordinates of target 1 are (x 1 ,y 1 ,z 1 ), the point cloud coordinates of target 2 are (x 2 ,y 2 ,z 2 ).

[0054] S500: Extracting feature data of the changed target according to the supplemented changed point cloud graph.

[0055] Among them, the feature data includes the target size, target position and target outline; according to the above-mentioned change area including target 1 and target 2, the size, position and outline of target 1 and target 2 are extracted, for example: the size of target 1 can be 1m×0.8m×2m, the position of target 1 relative to target 2, and the outline of target 1; the size of target 2 can be 1.5m×1m×1.8m, the position of target 2 relative to target 1, and the outline of target 2.

[0056] S600: If any item in the feature data of the changed target is different from that of the initial target, the changed target is added to the target list, and the three-dimensional laser radar array and the camera array are controlled to track the changed target to obtain the position and image of the changed target.

[0057] For example: if the initial targets are target 1 and target 2, and the changed target is target 3, if the size of target 3 is different from that of target 1 or target 2, target 3 is added to the target list; if the position of target 3 is different from that of target 1 or target 2, target 3 is added to the target list; if the outline of target 3 is different from that of target 1 or target 2, target 3 is also added to the target list, and then target 3 is tracked using a three-dimensional lidar array and a camera array to obtain the position and image of target 3. If the size, position and outline of target 3 are the same as those of target 1 or target 2, target 3 is not added to the target list.

[0058] The present embodiment provides a method for regional target detection and tracking, which can connect a three-frequency microwave radar, a three-dimensional laser radar, and a camera to a data processing module. Compared with a single microwave radar, the method is more responsive, accurately tracks targets, and displays more intuitively, with less interference between different sensors. The present embodiment only needs to point to the corresponding area when needed, and does not need to install a large number of three-dimensional laser radars to cover the blind spots of the area, so the response is faster; by measuring target information through radar, it is not easily affected by abnormal lighting conditions, and can accurately judge targets with similar colors and blurred outlines.

[0059] Before the three-frequency microwave radar array obtains the echo characteristics of the target area, it is necessary to initialize the three-frequency microwave radar array, the three-dimensional laser radar array, the camera array and the data processing module and pre-process the target area. Therefore, in some embodiments, the target point cloud coordinates of the target area can be scanned by the three-dimensional laser radar; then the target point cloud coordinates are spliced ​​into the same scene to generate an initial point cloud map of the target area; by identifying the targets in the initial point cloud map, a target list is formed, wherein the target list includes characteristic data of each target; and the echo characteristics of the target area are obtained by the three-frequency microwave radar array.

[0060] It is understandable that during the initialization process, there may be no target in the target area. Therefore, the target list formed by the initial point cloud image may be an empty list, that is, there is no target or target feature in the target list. Of course, there may be targets during the initialization process, and in this case, there are targets in the target list.

[0061] In order to prevent the three-dimensional laser radar and camera from wasting resources, in some embodiments, if all items in the feature data of the changed target are the same as those of the initial target, the three-dimensional laser radar and camera in the changed area will be set to an idle state. That is to say, when the feature data of the changed target is the same as that of the initial target, the data processing module will release the three-dimensional laser radar resources and the camera resources, so that the three-dimensional laser radar and the camera remain in an inoperative state, thereby achieving the effect of saving costs.

[0062] When scheduling an idle three-dimensional laser radar array, in order to make the point cloud map of the changed area clearer, in some embodiments, the method also includes: first obtaining the coverage area of ​​the three-dimensional laser radar, wherein the coverage area is divided into at least four categories, and the weight of each category of coverage area is different; then calculating the scheduling score of the three-dimensional laser radar, the scheduling score is the product of the weight of the coverage area and the area of ​​the coverage area; then setting the scheduling priority of the three-dimensional laser radar according to the scheduling score, and starting the three-dimensional laser radar according to the scheduling priority to complete the scheduling of the three-dimensional laser radar according to the scheduling score.

[0063] See also Fig.16 It can be seen that the coverage area can be divided into four categories: A, B, C, and D. Sensor No. 1 is a sensor in a 3D laser radar, and Sensor No. 2 is a sensor in another 3D laser radar. It can be seen that for both Sensor No. 1 and Sensor No. 2, when the sensor direction is toward Area A, the area with the clearest point cloud is Area A. Of course, if Sensor No. 1 is facing Area B, the area with the clearest point cloud may be Area B. Fig.16 For example, taking the direction of the sensor toward area A as an example, according to the weights of areas A, B, C, and D, the weight order is A>B>C>D, that is, when scheduling, area A is scheduled first, and the point cloud image formed by illuminating area A is also clearer.

[0064] Because in the scheduling process, it is first necessary to determine which sensors point to the required area. For example, the area to be illuminated is area A. At this time, there may be two idle sensors that can illuminate the target area at the same time. In order to save resources, it is necessary to select one of the sensors, which can not only illuminate the target area, but also make the illuminated point cloud map clearer. That is, by calculating the scheduling score of the 3D laser radar, the scheduling score is the product of the weight of the coverage area and the area of ​​the coverage area. For example, the weights of A, B, C, and D are 5, 4, 3, and 2 respectively, and the areas of A, B, C, and D are The area of ​​the target area is 9㎡. The illumination direction of sensor 1 and sensor 2 covers the target area. For sensor 1, the target area is area B, and the area covered by sensor 1 is 3㎡. For sensor 2, the target area is area A, and the area covered by sensor 2 is 4㎡. At this time, the scheduling score of sensor 1 is 4×3=12, and the scheduling score of sensor 2 is 5×4=20. According to the scheduling score, the scheduling score of sensor 2 is greater than that of sensor 1. Therefore, the 3D laser radar corresponding to sensor 2 is scheduled first. It can be understood that if multiple sensors have the same scheduling score, any 3D laser radar can be randomly selected for scheduling.

[0065] See also Figure 4 , Figure 5 , Figure 6 In some embodiments, the target area can be pre-set as a polygonal area. When the target area is not a polygonal area, the target area can be divided into multiple polygonal sub-areas according to the area of ​​the target area and the area of ​​the polygonal area, and the tracking range of the three-dimensional laser radar and the camera can be set according to the sub-areas.

[0066] For example, the target area proposed in this embodiment is a rectangle with an area of ​​20 square meters. Figure 4 In the example, the target area is divided into rectangular areas of area 1, area 2, ... area N. Area 1, area 2, ... area N are sub-areas, and the area of ​​each sub-area is 20 m2. Figure 5 If an area that does not conform to the rectangle is encountered, the target area can be segmented so that the segmented area is divided into several approximately rectangular areas, so that the area of ​​each area is equal; Figure 6 For areas with multiple interactive traffic directions, the interactive area can be cleared first, and a monitoring area can be set at the exit to make the area of ​​each area equal; it can be understood that there are other areas that do not conform to polygons, and the shape, scope, etc. of the area can be specified by the area.

[0067] In some embodiments, the initial position of the changing target can also be recorded; a multi-frame point cloud map of the target area is obtained through a three-dimensional laser radar array; the moving position of the changing target in the multi-frame point cloud map is traversed; the moving distance is calculated based on the moving position and the initial position, and the movement time is calculated based on the number of frames of the multi-frame point cloud map; finally, the movement speed of the changing target is calculated based on the moving distance and the movement time.

[0068] For changing targets, the three-dimensional lidar array and the camera can also be controlled to track the target, and then the moving speed of the changing target can be calculated. The moving position and the initial position of the changing target can be used to calculate the moving distance. For example, the moving distance is 0.2m, the frame interval time of the multi-frame point cloud image is 0.033s, and it moves 0.2m in 20 frames, that is, it takes 0.66s to move 0.2m, and the moving speed is 0.3m / s.

[0069] Based on the above-mentioned method for detecting and tracking regional targets, some embodiments of the present application also provide a system for detecting and tracking regional targets, including: a three-frequency microwave radar array, a data processing module, a three-dimensional laser radar array and a camera array, wherein the three-frequency microwave radar array is used to obtain the echo characteristic change state of the target area; the data processing module is used to: determine the changed area through the change state; splice the point cloud coordinates adjacent to the changed area in the initial point cloud map with the point cloud coordinates of the changed area to generate a supplementary changed point cloud map; extract feature data of the changed target according to the supplementary changed point cloud map; if any item in the feature data of the changed target is different from the initial target, add the changed target to the target list; control the three-dimensional laser radar array and the camera array to track the changed target to obtain the position and image of the changed target; the three-dimensional laser radar array is used to obtain the point cloud map of the changed area, and is also used to track the changed target to obtain the position of the changed target; the camera array is used to track the changed target to obtain the image of the changed target.

[0070] During installation, the three-frequency microwave radar array, three-dimensional laser radar array, camera array and data processing module are connected so that they can communicate with each other. During installation, it is necessary to measure the offset value and rotation value of each unit coordinate system in the three-frequency microwave radar array, three-dimensional laser radar array and camera array relative to the monitoring area coordinate system. For example: the monitoring area coordinate system uses the two sides of the rectangular area as the X-axis and the Y-axis, and the direction perpendicular to the ground as the Z-axis. The angle between the sensor's own coordinate system of the three-frequency microwave radar and the three-dimensional laser radar and the X-axis of the system coordinate system is P, and the angle between the sensor's own coordinate system and the Z-axis is Φ. The offset value (X, Y, Z) and rotation value (P, Φ) of each sensor unit will be input into the data processing module as parameters. The data processing module will unify the data coordinate system according to the offset value and rotation value of each sensor unit, unify all sensor output results to the global coordinate, and unify the coordinates of the three-frequency microwave radar array, three-dimensional laser radar array, camera array and data processing module.

[0071] Among them, the data processing module can adopt a microprogram controller, and the microprogram controller includes an instruction sequence, for example: the instruction sequence may include five microinstructions, including the first microinstruction for determining the changed area through the changed state, the second microinstruction for splicing the point cloud coordinates adjacent to the changed area in the initial point cloud map with the point cloud coordinates of the changed area to generate a supplementary changed point cloud map, the third microinstruction for extracting the feature data of the changed target according to the supplementary changed point cloud map, the fourth microinstruction for adding the changed target to the target list if any item in the feature data of the changed target is different from the initial target, and the fifth microinstruction for controlling the three-dimensional lidar array and the camera array to track the changed target to obtain the position and image of the changed target.

[0072] The data processing module can also adopt different microinstructions according to different requirements. For example: a new instruction is to set the three-dimensional laser radar and camera in the changed area to idle state if all items in the feature data of the changed target are the same as the initial target. Another example: a new instruction is to obtain the coverage area of ​​the three-dimensional laser radar; calculate the scheduling score of the three-dimensional laser radar; set the scheduling priority of the three-dimensional laser radar according to the scheduling score; start the three-dimensional laser radar according to the scheduling priority. It can be understood that microinstructions can also be added by expanding the microinstructions in the microprogram controller. Of course, the control requirements can also be modified by modifying the microinstructions.

[0073] In some embodiments, the three-frequency microwave radar array includes multiple three-frequency microwave radar units, and the three-frequency microwave radar units are arranged at the end points or inside the target area; the three-dimensional laser radar array includes multiple three-dimensional laser radars, and the three-dimensional laser radar includes at least one solid-state radar and at least one high-speed pan-tilt head, or includes at least one multi-line radar and at least one high-speed pan-tilt head, and the high-speed pan-tilt head is a two-dimensional radar; the camera array includes multiple cameras.

[0074] See also Figure 7 , for a three-frequency microwave radar array including Figure 7 The microwave radars of all frequencies in the three-frequency microwave radar array, frequency A, frequency B, and frequency C can be used as a three-frequency microwave radar unit. However, it should be noted that during installation, the three frequencies of the three-frequency microwave radar array need to be placed alternately so that two adjacent three-frequency microwave radar units can scan adjacent areas. The two adjacent three-frequency microwave radar units have different frequencies, so that when the three-frequency microwave radar is working, the three-frequency microwave radar units of the same frequency can be started alternately to reduce mutual interference.

[0075] The 3D LiDAR array consists of two LiDARs with different frequencies, which are used to scan the area in detail. Fig.10 Each unit in the array can be composed of a solid-state radar and a high-speed gimbal. The high-speed radar is a two-dimensional radar that supports pitch and yaw rotation, driving the solid-state radar to point to the area that needs to be scanned; Fig.11 Each unit in the array can also be replaced by a multi-line three-dimensional radar instead of a solid radar, that is, it consists of a multi-line radar and a high-speed pan-tilt head; see Fig.12 , without using high-speed radar, multi-line radar can also be used, but it should be understood that in this case, the number of multi-line radars needs to be adaptively increased.

[0076] During installation, see Fig.13 , the 3D laser radar array covers the entire monitoring area, where two 3D laser radars with different frequencies are installed back to back, so that the 3D laser radars with the same frequency can be reduced to irradiate each other and cause interference; see Fig.14 , the default illumination direction of the 3D laser radar is parallel to the main traffic direction of the area, see Figure 3 In this embodiment, the X direction is taken as the main direction of travel in the monitoring area by default, and the three-dimensional laser radar can swing left and right driven by the gimbal to achieve full coverage of the area.

[0077] When installing the camera array, see Fig.15 The camera array consists of multiple spherical network cameras. When installed, they need to evenly cover the entire monitoring area, and the default pointing direction of the camera is also parallel to the main traffic direction of the area.

[0078] It can be seen from the above technical solutions that this embodiment provides a method and system for regional target detection and tracking, which obtains the echo feature change state of the target area through a three-frequency microwave radar array; and determines the changed area through the change state; and obtains the point cloud map of the changed area through a three-dimensional laser radar array; then splices the point cloud coordinates adjacent to the changed area in the initial point cloud map with the point cloud coordinates of the changed area to generate a supplementary change point cloud map; then extracts the feature data of the changed target based on the supplementary change point cloud map, and the feature data includes the target size, target position and target outline; if any item in the feature data of the changed target is different from the initial target, the changed target is added to the target list, and the three-dimensional laser radar array and the camera array are controlled to track the changed target to obtain the position and image of the changed target. The method extracts the feature data of the changed target through the supplementary change point cloud map, and accurately identifies the changed target in combination with the initial target to solve the problem of being unable to accurately detect and track the targets in the monitoring area.

[0079] Similar parts between the embodiments provided in this application can be referenced to each other. The specific implementation methods provided above are only a few examples under the general concept of this application and do not constitute a limitation on the protection scope of this application. For those skilled in the art, any other implementation methods expanded based on the scheme of this application without creative work belong to the protection scope of this application.

Claims

1. A regional target detection and tracking method, It is characterized in that include: Acquire the echo characteristic change state of the target area through a three-frequency microwave radar array; Determine a change region through the change state, wherein the change region includes a plurality of change targets; Acquire a point cloud map of the change area through a three-dimensional laser radar array, wherein the point cloud map includes a plurality of point cloud coordinates; splicing the point cloud coordinates adjacent to the changed area in the initial point cloud map with the point cloud coordinates of the changed area to generate a supplemented changed point cloud map, wherein the initial point cloud map includes a target list, and the target list includes a plurality of initial targets; Extracting feature data of the changed target according to the supplemented change point cloud image, wherein the feature data includes target size, target position and target outline; If any item in the characteristic data of the changed target is different from the initial target, the changed target is added to the target list, and the three-dimensional lidar array and the camera array are controlled to track the changed target to obtain the position and image of the changed target.

2. The method for regional target detection and tracking according to claim 1, It is characterized in that The method further comprises: Scanning the target point cloud coordinates of the target area by the three-dimensional laser radar, wherein the three-dimensional laser radar array includes a plurality of three-dimensional laser radars; The target point cloud coordinates are stitched into the same scene to generate an initial point cloud map of the target area; Identifying objects in the initial point cloud image to form an object list, wherein the object list includes the feature data of each object; The echo characteristics of the target area are acquired through a three-frequency microwave radar array.

3. The method for regional target detection and tracking according to claim 2, It is characterized in that The method further comprises: Performing environmental filtering on the echo feature to obtain a filtered echo feature; The filtered echo characteristics are compared with the initial echo characteristics to determine a change state of the filtered echo characteristics.

4. The method for regional target detection and tracking according to claim 2, It is characterized in that The method further comprises: If all items in the feature data of the changed target are the same as those of the initial target, the three-dimensional laser radar and the camera in the changed area are set to an idle state.

5. The method for regional target detection and tracking according to claim 2, It is characterized in that The method further comprises: Acquire the coverage area of ​​the three-dimensional laser radar, where the coverage area is divided into at least four categories, and each category of the coverage area has a different weight; Calculating a dispatch score of the three-dimensional laser radar, where the dispatch score is the product of a weight of the coverage area and an area of ​​the coverage area; The scheduling priority of the three-dimensional laser radar is set according to the scheduling score, and the three-dimensional laser radar is started according to the scheduling priority.

6. The method for regional target detection and tracking according to claim 5, It is characterized in that The target area is a polygonal area; the method further comprises: If the target area is not a polygonal area, dividing the target area into a plurality of polygonal sub-areas according to the area of ​​the target area and the area of ​​the polygonal area; The tracking range of the three-dimensional laser radar and the camera is set according to the sub-area.

7. The method for regional target detection and tracking according to claim 1, It is characterized in that The method further comprises: If the change state is a change, determine the change area; According to the ID of the three-frequency microwave radar unit that outputs the change state, a change area monitored by the three-frequency microwave radar unit is determined, and the three-frequency microwave radar array includes a plurality of three-frequency microwave radar units.

8. The method for regional target detection and tracking according to claim 7, It is characterized in that The method further comprises: Recording the initial position of the changed target; Acquire a multi-frame point cloud image of the target area through the three-dimensional laser radar array; Traversing the moving positions of the changed targets in the multi-frame point cloud images; Calculating a moving distance according to the moving position and the initial position, and calculating a moving time according to the number of frames of the multi-frame point cloud image; The moving speed of the changing target is calculated according to the moving distance and the moving time.

9. A regional target detection and tracking system, It is characterized in that include: A three-frequency microwave radar array is used to obtain the echo characteristic change status of the target area; Data processing module for: determining a change region according to the change state; splicing the point cloud coordinates adjacent to the changed area in the initial point cloud map with the point cloud coordinates of the changed area to generate a supplemented changed point cloud map; Extracting feature data of the change target according to the supplemented change point cloud map; If any item of the characteristic data of the changed target is different from that of the initial target, the changed target is added to the target list; Controlling the three-dimensional laser radar array and the camera array to track the changing target to obtain the position and image of the changing target; The three-dimensional laser radar array is used to obtain a point cloud image of the changed area and is also used to track the changed target to obtain the position of the changed target; The camera array is used to track the changing target to obtain an image of the changing target.

10. The regional target detection and tracking system according to claim 9, It is characterized in that include: The three-frequency microwave radar array includes a plurality of three-frequency microwave radar units, and the three-frequency microwave radar units are arranged at the end points or inside of the target area; The three-dimensional laser radar array includes a plurality of three-dimensional laser radars, and the three-dimensional laser radar includes at least one solid-state radar and at least one high-speed pan-tilt, or includes at least one multi-line radar and at least one high-speed pan-tilt, and the high-speed pan-tilt is a two-dimensional radar; The camera array includes a plurality of cameras.

Images