A low, slow and small target detection system and method based on laser radar

Abstract

The application provides a low, slow and small target detection system based on a laser radar, which comprises a power supply subsystem, a control subsystem, a servo turntable, a laser radar subsystem and a data processing unit; the servo turntable provides a rotating platform for the laser radar subsystem, so that the laser radar subsystem rotates at a constant speed in a horizontal direction by 360 degrees; the control subsystem is connected with the servo turntable, so as to realize zero return, speed control and position information reading of the servo turntable; the power supply subsystem supplies power to the control subsystem, the servo turntable, the laser radar subsystem and the data processing unit; the laser radar subsystem is used for acquiring point cloud data in a scanning range; and the data processing unit processes the point cloud data, so as to realize detection of a low, slow and small target. The application is not affected by background environment light, electromagnetic radiation and noise, has strong anti-interference ability, has high target three-dimensional space coordinate and size measurement precision, and has low false alarm rate and low missing alarm rate.

Description

Technical Field

[0001] This invention belongs to the field of radar detection, specifically relating to a system and method for detecting low, slow, and small targets based on lidar. Background Technology

[0002] In recent years, the number of low-altitude, slow-moving, and small (LASM) targets, primarily drones, has increased dramatically, posing safety hazards to various sectors of society. Existing LASM detection systems mainly include: traditional radar detection systems, visible light imaging detection systems, thermal infrared imaging detection systems, radio frequency (RF) detection systems, acoustic signature detection systems, and combinations of these technologies. However, the small radar cross-section of LASM targets, the low signal-to-noise ratio of radar echoes, and interference from electromagnetic clutter in lower altitudes make it difficult for traditional radar to fully utilize its advantages in detecting LASM targets. Visible light imaging detection systems cannot operate at night. Since LASM targets are primarily powered by electricity and emit relatively little heat, the detection capability of thermal infrared imaging detection systems is limited. RF and acoustic signature detection systems detect LASM targets by detecting the radio and acoustic signals emitted by them, but these are easily affected by environmental noise and have low positioning accuracy. Summary of the Invention

[0003] This invention takes into account the advantages of lidar active detection, which is unaffected by background ambient light, electromagnetic radiation, and noise, has strong anti-interference capabilities, and offers high ranging accuracy and angular resolution. It can acquire high spatial resolution and high-precision 3D point cloud data, providing sufficient environmental information to distinguish low-speed, small targets from other targets and buildings. This invention proposes a lidar-based low-speed, small target detection system and method, improving the anti-interference capability and detection accuracy of existing detection methods.

[0004] The technical solution of the present invention is as follows:

[0005] A low-speed, small target detection system based on lidar, the system comprising: a power supply subsystem, a control subsystem, a servo turntable, a lidar subsystem, and a data processing unit;

[0006] The servo turntable provides a rotating platform for the lidar subsystem, enabling the lidar subsystem to rotate 360° at a constant speed in the horizontal direction.

[0007] The control subsystem is connected to the servo turntable to realize servo turntable zeroing, speed control, and position information reading;

[0008] The power supply subsystem provides power to the control subsystem, servo turntable, lidar subsystem and data processing unit;

[0009] The lidar subsystem is used to acquire point cloud data within the scanning range;

[0010] The data processing unit processes the point cloud data to achieve the detection of low, slow, and small targets.

[0011] Furthermore, the system also includes host computer software; the host computer software runs on the data processing unit, on the one hand realizing servo turntable control and position information reading, and on the other hand, synchronizing the time of the lidar in the lidar subsystem, controlling the lidar working mode, reading the point cloud data collected by the lidar, processing the point cloud data, and displaying the processing results in real time.

[0012] Furthermore, the lidar subsystem consists of a line-scan lidar, a network switch, and a slip ring;

[0013] The power supply subsystem supplies power to the line-scan lidar and network switch via slip rings; the line-scan lidar is connected to the data processing unit via the network switch and slip rings.

[0014] Furthermore, the line-scanning lidar performs uniform line scanning within a 90° range in the vertical direction and rotates 360° in the horizontal direction to acquire point cloud data in the hemispherical space.

[0015] Furthermore, the line-scanning lidar is configured to consist of three units.

[0016] Furthermore, the included angle between the line-scanning lidars is 120°.

[0017] The present invention also provides a method for detecting small, slow, low-altitude targets based on lidar, the method comprising the following steps:

[0018] S1 uses a precise time protocol to synchronize the time of three line-scan lidars, ensuring that the point cloud data of the three lidars have the same time information.

[0019] S2 Click the radar connection button to connect the lidar and initialize the lidar;

[0020] S3 Click the turntable connection button to connect the turntable, set the turntable speed, and send a zero-return command to the turntable. The turntable will then perform the zero-return operation.

[0021] S4 clicks the Start Detection button. The data processing unit acquires the lidar data and the position data of the servo turntable in real time. Based on the time information of the point cloud data of the three lidars and the position information of the servo turntable, the point cloud data of the three lidars is converted into the world coordinate system of the servo turntable, thus unifying the coordinate system of the point cloud data of the three lidars. The lidar point cloud data is then processed in real time to determine whether there are low, slow and small targets in the airspace to be measured.

[0022] Furthermore, S4 specifically includes:

[0023] Based on the time information of the point cloud data from the three lidars and the position information of the turntable, the point cloud data of the three lidars is converted into the world coordinate system of the servo turntable, thereby unifying the coordinate system of the point cloud data of the three lidars and obtaining the point cloud data of the hemispherical space of the monitoring area.

[0024] Construct an octree tree1 for the background frame point cloud that does not contain low-speed, small targets; construct an octree tree2 for the current frame point cloud based on the point cloud data of the hemispherical space of the monitoring area.

[0025] Calculate the difference between tree2 and tree1, and consider the nodes of tree2-tree1 as foreground objects, and the remaining nodes as background objects. Denote the set of all points in the octree nodes of the foreground objects as P1, and use the nearest neighbor algorithm to calculate the nearest distance D from all points in point set P1 to the center of the octree node in the background frame. min Using the voxel side length R of the octagonal leaf node as the judgment threshold, the following conditions will be met. The points identified as misjudged background points are considered to be foreground points, and the background points are removed.

[0026] For any unvisited point p in the foreground points i (x i ,y i ,z i Determine the Eps parameter for that point. β represents the average angular resolution of the lidar, and a minimum number of points (Minpts) is set. Next, the number N of all points whose distance from this point is less than or equal to Eps is calculated. Eps If N Eps If the value is greater than or equal to Minpts, then the point is considered the core point, and a data cluster C is established, including the point and all its neighboring points. Next, another unvisited data point is selected in cluster C, and the above steps are repeated until no new data points are added to cluster C. Finally, the above steps are repeated until all points are classified. Each class corresponds to a low-slow-small objective.

[0027] If a low-speed, small target is present, its location information will be displayed in real time on the host computer software; if no low-speed, small target is present, the LiDAR data will continue to be acquired in real time, and the low-speed, small target detection operation will be performed.

[0028] Technical effects of the present invention:

[0029] The advantages of the low-speed-small detection system based on lidar of the present invention are: it is not affected by background ambient light, electromagnetic radiation and noise, and has strong anti-interference ability; it has high accuracy in measuring the three-dimensional spatial coordinates and size of the target; and it has low false alarm rate and low missed alarm rate. Attached Figure Description

[0030] The accompanying drawings illustrate various embodiments generally by way of example rather than limitation, and are used, together with the specification and claims, to explain embodiments of the invention. Where appropriate, the same reference numerals are used in all drawings to refer to the same or similar parts. Such embodiments are illustrative and are not intended to be exhaustive or exclusive embodiments of the apparatus or method.

[0031] Figure 1 A schematic diagram of the electrical connection of the system according to the present invention is shown;

[0032] Figure 2 A schematic diagram of the system mechanical structure of the present invention is shown;

[0033] Figure 3 A schematic diagram of the working process of the detection system of the present invention is shown;

[0034] Figure 4 The host computer software of the present invention is shown;

[0035] Reference numerals: 1-First line scanning lidar, 2-LiDAR network interface, 3-LiDAR power interface, 4-Second line scanning lidar, 5-Support plate, 6-Slip ring power interface, 7-Slip ring network interface, 8-Slip ring, 9-Servo turntable, 10-Control and feedback line interface, 11-Turntable power interface, 12-Slip ring network interface and power interface, 13-Conversion board, 14-Third line scanning lidar, 15-Network switch. Detailed Implementation

[0036] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0037] This invention provides a low-speed, small target detection system based on lidar; such as Figure 1 As shown, the system consists of a power supply subsystem, a servo turntable, a lidar subsystem, a data processing unit, and host computer software.

[0038] The power supply subsystem converts 220V AC to 24V DC to power the control subsystem, which in turn outputs the 24V DC to power the servo turntable. The power supply subsystem also converts 220V AC to 12V DC, which powers the lidar subsystem via a slip ring fixed to the servo turntable. Finally, the power supply subsystem converts 220V AC to 20V DC to power the data processing unit.

[0039] The lidar subsystem consists of a network switch and three line-scan lidars. The line-scan lidars are connected to the network switch via Ethernet cables. The Ethernet switch uses slip rings and Ethernet cables to enable communication between the line-scan lidars and the data processing unit.

[0040] The data processing unit is connected to the control subsystem via an Ethernet cable. The host computer software running on the data processing unit sends servo turntable control commands to the control subsystem and reads the servo turntable's position information. The control subsystem and the servo turntable are connected via control lines, position feedback lines, and a DC24V power supply line, enabling servo turntable zeroing, speed control, and position information reading. The data processing unit is connected to the network switch in the lidar subsystem via a slip ring and an Ethernet cable. The host computer software configures the parameters of the line-scan lidar in the lidar subsystem, reads lidar data, and processes the lidar data to achieve low-altitude, slow-moving, small target detection.

[0041] like Figure 2 As shown, three line-scan LiDARs are fixed to the adapter plate via support plate 5, with an included angle of 120° between them. The network interfaces of the three line-scan LiDARs are connected to the network interfaces on the network switch. The network switch 15 is also fixed to the adapter plate with screws. The adapter plate 13 is fixed to the servo turntable 9 with screws, enabling the servo turntable 9 to drive the line-scan LiDARs to rotate 360°. A slip ring 8 is installed in the middle of the servo turntable. The upper end of the slip ring 8 includes a slip ring network cable interface and a power interface. The network interface is connected to the network interface on the network switch 15, and the power interface is connected to the power interfaces of the network switch and the three line-scan LiDARs. The network interface and power interface at the lower end of the slip ring 8 are led out through the servo turntable base. The power interface 11 and control and feedback line interface 10 of the servo turntable 9 are also located on the turntable base.

[0042] Figure 3 To illustrate the system workflow, before the system starts operating, open the host computer software running on the data processing unit. Figure 4 Create the host computer software interface and perform the following operations:

[0043] S1 uses a precise time protocol to synchronize the time of three line-scan lidars, ensuring that the point cloud data of the three lidars have the same time information.

[0044] S2 Click the radar connection button to connect the lidar and initialize the lidar;

[0045] S3 Click the turntable connection button to connect the turntable, set the turntable speed, and send a zero-return command to the turntable. The turntable will then perform the zero-return operation.

[0046] S4 clicks the Start Detection button. The data processing unit acquires the lidar data and the position data of the servo turntable in real time. Based on the time information of the point cloud data of the three lidars and the position information of the servo turntable, the point cloud data of the three lidars is converted into the world coordinate system of the servo turntable, thus unifying the coordinate system of the point cloud data of the three lidars. The lidar point cloud data is then processed in real time to determine whether there are low, slow and small targets in the airspace to be measured.

[0047] Specifically, S4 is:

[0048] Based on the time information of the point cloud data from the three lidars and the position information of the turntable, the point cloud data of the three lidars is converted into the world coordinate system of the servo turntable, thereby unifying the coordinate system of the point cloud data of the three lidars and obtaining the point cloud data of the hemispherical space of the monitoring area.

[0049] Construct an octree tree1 for the background frame point cloud that does not contain low-speed, small targets; construct an octree tree2 for the current frame point cloud based on the point cloud data of the hemispherical space of the monitoring area.

[0050] Calculate the difference between tree2 and tree1, and consider the nodes of tree2-tree1 as foreground objects, and the remaining nodes as background objects. Denote the set of all points in the octree nodes of the foreground objects as P1, and use the nearest neighbor algorithm to calculate the nearest distance D from all points in point set P1 to the center of the octree node in the background frame. min Using the voxel side length R of the octagonal leaf node as the judgment threshold, the following conditions will be met. The points identified were considered misjudged as background points, while the remaining points were considered as foreground points, and the background points were removed.

[0051] For any unvisited point p in the foreground points i (x i ,y i ,z i Determine the Eps parameter for that point. β represents the average angular resolution of the lidar, and a minimum number of points (Minpts) is set. Next, the number N of all points whose distance from this point is less than or equal to Eps is calculated. Eps If N Eps If the value is greater than or equal to Minpts, then the point is considered the core point, and a data cluster C is established, which includes the point and all its neighboring points. Next, another unvisited data point is selected in cluster C, and the above steps are repeated until no new data points are added to cluster C. Finally, the above steps are repeated until all points are classified. Each class corresponds to a low-slow-small target, and the points in the class are the points on the low-slow-small target.

[0052] If a low-speed, small target is present, its location information will be displayed in real time on the host computer software; if no low-speed, small target is present, the LiDAR data will continue to be acquired in real time, and the low-speed, small target detection operation will be performed.

[0053] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the technical scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for detecting small, slow-moving targets based on lidar, characterized in that, The method includes the following steps: S1 uses a precise time protocol to synchronize the time of three line-scan lidars, ensuring that the point cloud data of the three lidars have the same time information. S2 Click the radar connection button to connect the lidar and initialize the lidar; S3 Click the turntable connection button to connect the turntable, set the turntable speed, and send a zero-return command to the turntable. The turntable will then perform the zero-return operation. S4 Click the Start Detection button. The data processing unit acquires the lidar data and the position data of the servo turntable in real time. Based on the time information of the point cloud data of the three lidars and the position information of the servo turntable, the point cloud data of the three lidars is converted into the world coordinate system of the servo turntable, thus unifying the coordinate system of the point cloud data of the three lidars and obtaining the point cloud data of the hemispherical space of the monitoring area. The point cloud data is then processed in real time to determine whether there are low, slow and small targets in the airspace to be measured. Specifically, S4 is: Based on the time information of the point cloud data from the three lidars and the position information of the turntable, the point cloud data of the three lidars is converted into the world coordinate system of the servo turntable, thereby unifying the coordinate system of the point cloud data of the three lidars and obtaining the point cloud data of the hemispherical space of the monitoring area. Construct an octree tree1 for the background frame point cloud that does not contain low-speed, small targets; construct an octree tree2 for the current frame point cloud based on the point cloud data of the hemispherical space of the monitoring area. Calculate the difference between tree2 and tree1, considering the nodes of tree2-tree1 as foreground elements and the remaining nodes as background elements; denote the set of all points in the octree nodes of the foreground elements as... The nearest neighbor algorithm is used to calculate the point set. The closest distance of all points from the center of the octree node in the background frame Using the voxel side length R of the octagonal leaf node as the judgment threshold, the following conditions will be met. The points identified as misjudged background points are considered to be foreground points, and the background points are removed. For any unvisited point in the foreground. Determine about this point parameter, , Set the average angular resolution of the lidar and the minimum number of points. ; Secondly, calculate the distance from this point to be less than or equal to Number of all points ,like If a point is identified as the core point, a data cluster C is created, including the point and all its neighboring points. Then, another unvisited data point is selected in cluster C, and the above steps are repeated until no new data points are added to cluster C. Finally, the above steps are repeated until all points are classified. Each class corresponds to a low-slow-small objective.

2. The detection method according to claim 1, characterized in that, The method further includes: If a low-speed, small target is present, its location information will be displayed in real time on the host computer software; if no low-speed, small target is present, the LiDAR data will continue to be acquired in real time, and the low-speed, small target detection operation will be performed.

3. A low-altitude, slow-moving, small target detection system based on lidar, characterized in that, The system uses the low-speed, small target detection method based on lidar as described in any one of claims 1-2, and the system includes: a power supply subsystem, a control subsystem, a servo turntable, a lidar subsystem, and a data processing unit; The servo turntable provides a rotating platform for the lidar subsystem, enabling the lidar subsystem to rotate 360° at a constant speed in the horizontal direction. The control subsystem is connected to the servo turntable to realize servo turntable zeroing, speed control, and position information reading; The power supply subsystem provides power to the control subsystem, servo turntable, lidar subsystem and data processing unit; The lidar subsystem is used to acquire point cloud data within the scanning range; The data processing unit processes the point cloud data to achieve the detection of low, slow, and small targets.

4. The detection system according to claim 3, characterized in that, The system also includes host computer software; the host computer software runs on the data processing unit, on the one hand realizing servo turntable control and position information reading, and on the other hand, synchronizing the time of the lidar in the lidar subsystem, controlling the lidar working mode, reading the point cloud data collected by the lidar, processing the point cloud data, and displaying the processing results in real time.

5. The detection system according to claim 3, characterized in that, The lidar subsystem consists of a line-scan lidar, a network switch, and a slip ring. The power supply subsystem supplies power to the line-scan lidar and network switch via slip rings; the line-scan lidar is connected to the data processing unit via the network switch and slip rings.

6. The detection system according to claim 5, characterized in that, The line-scanning lidar performs uniform line scanning within a 90° vertical range on one hand, and rotates 360° horizontally on the other hand to acquire point cloud data in a hemispherical space.

7. The detection system according to claim 5, characterized in that, The line-scan lidar is configured to consist of 3 units.

8. The detection system according to claim 5, characterized in that, The included angle between the line-scan lidars is 120°.

Images