Fusion processing method for improving angle estimation precision

Abstract

The invention provides a fusion processing method for improving angle estimation precision, and the method comprises the steps: 1, carrying out the same-address deployment of a radar and a photoelectric camera, and carrying out the deployment in a horizontal manner or a vertical manner; step 2, the radar performs rapid search on the whole airspace, and estimates the distance, the horizontal angle and the vertical angle of the target after the target is detected; 3, after receiving the coarse angle sum, the photoelectric camera rapidly rotates, adjusts the pointing direction of the optical axis, captures a target and performs imaging; 4, calculating a target deviation angle; comprising vertical angle deviation and horizontal angle deviation; and step 5, after obtaining the vertical angle deviation and the horizontal angle deviation of the target, correcting the target angle to obtain the accurate angle of the target. The method combines the advantages of the radar and the photoelectric camera, improves the estimation precision of the target angle, achieves the high-precision positioning of an unmanned plane target, and meets the control requirements of a low-altitude unmanned plane.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of low-altitude target detection and specifically relates to a radar and camera fusion angle estimation method, and particularly relates to a fusion processing method for improving angle estimation precision. BACKGROUND

[0002] With the development of low-altitude economy, the application of low-altitude unmanned aerial vehicles (UAVs) is rapidly growing, and the phenomenon of “black flight” and “random flight” occurs frequently, which seriously affects the order of low-altitude flight and even public security, and it is urgent to effectively control low-altitude UAVs. Low-altitude UAV target detection and parameter estimation are prerequisite conditions for the control of low-altitude UAVs.

[0003] The UAV target is a typical small target, and common low-altitude detection means includes radar, photoelectric camera, radio detection equipment, etc. The radar has a long detection distance and high search efficiency, and can measure the distance, speed, angle and other parameters of the target, but is limited by volume, weight, cost and processing real-time, and has low angle resolution and angle measurement precision. The angle measurement precision of a general low-altitude detection radar is at the level of 1°. Assuming that the azimuth / elevation angle measurement precision of the radar is 1° and the distance of the target is 1000 meters, the azimuth / elevation positioning error is close to 20 meters, which cannot meet the positioning precision requirements of the UAV in low-altitude detection. The photoelectric camera can image the target and has high resolution, and is indispensable in low-altitude target perception, but cannot accurately estimate the distance, speed and other parameters of the target, and the field of view is very small in long-distance detection, and the spatial search efficiency is low. The UAV is mobile and flexible, and the flight speed can reach more than 20 meters per second, and it is difficult to complete the detection and positioning of the high-maneuvering UAV in the full airspace only by using the photoelectric camera. The commonly used radio detection equipment includes passive radio positioning and reverse protocol analysis. The passive radio positioning includes AOA and TDOA measurement methods, and needs multi-station cooperative detection, which has great difficulty in station deployment, high deployment cost, poor ranging and lateral precision, and cannot meet the positioning precision requirements of low-altitude detection. The reverse protocol analysis obtains the position, speed and other information of the target by analyzing the communication protocol of the UAV, and has high positioning precision, but is only applicable to cooperative UAVs with open communication protocol, and is invalid for UAVs with encrypted communication protocol or private communication protocol and radio-silent UAVs.

[0004] For low-altitude UAV target detection and localization, radar is highly efficient for airspace reconnaissance, but its angle estimation accuracy is generally poor. Traditional methods for improving radar angle estimation accuracy are costly in terms of both hardware and computation. Electro-optical cameras play an irreplaceable role in low-altitude detection due to their high-resolution imaging capabilities, but their airspace search efficiency is relatively low. Neither a single radar nor an electro-optical camera can meet the needs of low-altitude UAV detection and localization. This invention utilizes the advantages of radar and camera airspace reconnaissance, and through fusion processing, improves the angle estimation accuracy for typical small targets like UAVs, thereby improving target localization accuracy and meeting the needs of low-altitude detection and early warning. Summary of the Invention

[0005] This invention proposes a fusion processing method to improve the accuracy of angle estimation. This method enhances target angle estimation accuracy, is easy to deploy, has a simple processing procedure, and ensures real-time detection and positioning. This novel method fully utilizes the advantages of radar and electro-optical cameras, obtaining high-precision angle information through joint processing. First, radar's rapid search capability scans the entire airspace to detect the target and obtain its distance and angle information. Since radar's angle measurement accuracy is relatively poor, this angle information is referred to as the coarse angle. The electro-optical camera adjusts its optical axis direction based on the radar's coarse angle information to acquire the target. Then, based on the target's position in the image plane, the target's offset angle relative to the optical axis is calculated. Combined with the coarse angle, the precise angle information of the target can be obtained.

[0006] This invention provides a fusion processing method to improve the accuracy of angle estimation, such as... Figure 1 As shown, the steps are as follows: Step 1: To simplify calculations, the radar and electro-optical camera are deployed at the same location. This can be done horizontally or vertically, as shown in the following deployment method: Figure 2a and Figure 2b As shown. Deployed horizontally, the center of the radar antenna and the center of the electro-optical camera lens are on the same horizontal plane, with a distance of [missing information]. Deployed vertically, the center of the radar antenna and the center of the electro-optical camera lens are on the same vertical plane, with a distance between the two centers of [missing information]. Center-to-center distance Keep it as small as possible, so as not to affect the rotation of the radar and camera. The presence of this feature causes the radar and photoelectric camera to be unable to be perfectly aligned in a certain dimension, resulting in a horizontal distance. This will cause horizontal angle errors and vertical spacing errors. It will introduce vertical angle errors, but for long-distance measurements, the errors are small. The resulting angular error is very small and can generally be ignored.

[0007] Step 2: The radar performs a rapid search of the entire airspace, and estimates the target's distance after detecting it. Horizontal angle and vertical angle Information such as... The parameter estimation here can employ any radar parameter estimation method, which is outside the scope of this invention and will therefore not be elaborated upon. and This is the target's coarse angle information; the radar sends the coarse angle to the electro-optical camera.

[0008] Step 3: The photoelectric camera receives the coarse angle. and Then, rotate quickly to adjust the direction of the optical axis. Orientation, target acquisition, and imaging. In the image plane, the image center location. The angle is Corresponding to the optical axis direction, the target center pixel Position coordinates are , for relative pixel offset, and Positive and negative basis The quadrant in the imaging plane coordinate system is determined. The positional relationship between the target and the optical axis is as follows: Figure 3 As shown.

[0009] Step 4: Calculate the target offset angle. The geometric relationship of the vertical offset angle is as follows: Figure 4 As shown, the imaging plane intersects the optical axis at... Point, the center of the camera lens is The camera focus is The camera focal length is Target center point The object distance is The center pixel of the target image The image distance is , arrive The distance is the image height offset vertical angle The angle between the optical axis and the horizontal plane. The angle offset is in the vertical direction of the target. This is based on the imaging geometry. ,in, , ; , for offset in the vertical direction distance, This refers to the vertical pixel size. The target's horizontal offset angle is also specified. ,in, , ; , for Horizontal offset distance, This refers to the pixel size in the horizontal direction.

[0010] Step 5: After obtaining the target's vertical and horizontal angle offsets, the target angles can be corrected to obtain the target's precise angles, including the precise vertical angle. Precise horizontal angle .

[0011] According to precise angle and radar measurement distance This allows us to obtain the target's precise location information. Since the error in calculating the angle offset on the imaging plane is at the pixel level, it can be completely ignored. Therefore, the corrected angle estimate can be considered to be very close to the true value, greatly improving the estimation accuracy without significantly reducing detection efficiency, thus meeting the needs of low-altitude UAV detection and positioning.

[0012] The advantages and beneficial effects of this invention are as follows: This invention proposes a fusion processing method to improve angle estimation accuracy, combining the respective advantages of radar and electro-optical cameras to improve the accuracy of target angle estimation, thereby achieving high-precision positioning of UAV targets and meeting the requirements for low-altitude UAV management. The fusion of radar and electro-optical cameras offers a large detection and monitoring range, high search efficiency, relatively easy joint deployment, and controllable cost, making it suitable for low-altitude security monitoring applications in important areas. The method proposed in this invention is simple to implement, has high detection efficiency, and can achieve high-precision angle measurement without increasing costs, facilitating engineering applications. Attached Figure Description

[0013] Figure 1 This is a flowchart of the radar-camera fusion method.

[0014] Figure 2a , Figure 2b This is a schematic diagram of radar and camera deployment methods.

[0015] Figure 3 This is a schematic diagram of the imaging plane coordinates.

[0016] Figure 4 This is a schematic diagram of the geometric relationship of the image plane. Detailed Implementation

[0017] To demonstrate the performance of the proposed method, a low-cost low-altitude detection radar and an electro-optical camera were used for verification. The radar angle estimation accuracy was 1°, and the distance estimation accuracy was 10 meters. The electro-optical camera was a visible-infrared dual-mode camera. The infrared camera was used for UAV observation, with a maximum focal length of 150 mm, a resolution of 640×512, and a pixel size of 12×12 micrometers. The observable target size for the UAV was approximately 0.4×0.4 meters. Based on a focal length of 150 mm, the field of view of the infrared camera was approximately 3°×2.3°, which could cover the target detected by the radar. The positioning coordinates of the UAV's onboard high-precision (centimeter-level) BeiDou real-time dynamic RTK (Real-time kinematic) positioning module were used as the true values ​​for the UAV's distance and angle. The UAV flew back and forth at a constant speed within a distance range of 990-1010 meters in a horizontal angle of 22° and a vertical angle of 6°.

[0018] Therefore, this invention proposes a fusion processing method to improve the accuracy of angle estimation, the specific implementation steps of which are as follows: Step 1: The radar and photoelectric camera are deployed horizontally, with the center of the radar antenna and the center of the photoelectric camera lens on the same horizontal plane, and the distance between the two centers is [missing information]. =0.5 meters. The presence of this element causes the radar and electro-optical camera to be unable to be perfectly aligned horizontally, resulting in a distance-dependent angular error. For targets at a distance of 1000 meters or more, this error is caused by… The resulting angular error is less than 0.03 degrees, which can generally be ignored. Since low-altitude detection typically requires a detection range of at least 1000 meters, the angular error caused by this is negligible. This results in angular errors.

[0019] Step 2: The radar performs a rapid search of the entire airspace, and estimates the target's distance after detecting it. =1010 meters, horizontal angle =21°, vertical angle =5°. and This refers to the target's coarse angle information; the radar will then interpret the coarse angle... Send to the photoelectric camera.

[0020] Step 3: The photoelectric camera receives the coarse angle. and Then, rotate quickly to adjust the direction of the optical axis. Orientation, target acquisition, and imaging. In the image plane, the image center location. The angle is The target center pixel is obtained by image detection, corresponding to the optical axis direction. Position coordinates are , and Positive and negative basis The quadrant in the imaging plane coordinate system is determined, where, =240 pixels, =207 pixels.

[0021] Step 4: Calculate the target offset angle. The geometric relationship of the vertical offset angle is as follows: Figure 4 As shown, the imaging plane intersects the optical axis at... Point, the center of the camera lens is The camera focus is The camera focal length is =150 mm, target center point The object distance is =1010 meters. Target image center pixel. Image distance is = (0.15×10¹⁰) / (10¹⁰-0.15)≈150.002 mm, which is approximately the focal length. arrive The distance is the image height offset =207×12e-3=2.484 mm. The vertical angular offset of the target is determined by the imaging geometry. = ≈0.95°. For the horizontal offset direction, the image distance is the same as in the vertical direction, which is... ≈150.002 mm, like width offset =240×12e-3=2.88 mm, target horizontal offset angle = ≈1.1°.

[0022] Step 5: After obtaining the target's vertical and horizontal angle offsets, the target angles can be corrected to obtain the target's precise angles, including the precise vertical angle. =5 + 0.95 = 5.95°, precise horizontal angle =21+1.1=22.1°.

[0023] According to precise angle and radar measurement distance This allows for the acquisition of precise target location information. Since the error in angle offset calculation on the imaging plane is at the pixel level and can be completely ignored, the corrected angle can be considered very close to the true angle value, greatly improving angle estimation accuracy. In this embodiment, the horizontal angle accuracy is improved from 1° to 0.1°, and the vertical angle accuracy is improved from 1° to 0.05°. For an infrared camera with a resolution of 640×512 and a pixel size of 12×12 micrometers, at a distance of 1000 meters, it can improve the horizontal and vertical angle accuracies by a maximum of 1.5° and 1.1° respectively (removing 6 pixels from the image edges to ensure target detection). In this test, after radar and camera fusion processing, the angle estimation accuracy of the UAV target can reach approximately 0.1°, which is 10 times higher than conventional radar angle estimation without reducing detection efficiency, meeting the needs of low-altitude UAV detection and positioning.

Claims

1. A fusion processing method for improving the accuracy of angle estimation, characterized in that, The steps are as follows: Step 1: Deploy radar and electro-optical cameras at the same location, either horizontally or vertically. Step 2: The radar performs a rapid search of the entire airspace, and estimates the target's distance after detecting it. Horizontal angle and vertical angle ; Step 3: The photoelectric camera receives the coarse angle. and Then, rotate quickly to adjust the direction of the optical axis. Direction, target acquisition and imaging; Step 4: Calculate the target offset angle; This includes vertical angle offset and horizontal angle offset; Step 5: After obtaining the target's vertical and horizontal angle offsets, correct the target angles to obtain the target's precise angles, including the precise vertical angle. Precise horizontal angle .

2. The fusion processing method for improving angle estimation accuracy according to claim 1, characterized in that: In step 1, the radar antenna and the photoelectric camera lens are deployed horizontally, with the center of the radar antenna and the center of the photoelectric camera lens on the same horizontal plane, and the distance between the two centers is [missing information]. .

3. The fusion processing method for improving angle estimation accuracy according to claim 1, characterized in that: In step 1, the radar antenna and the photoelectric camera lens are deployed vertically, with the center of the radar antenna and the center of the photoelectric camera lens on the same vertical plane, and the distance between the two centers is [missing information]. .

4. A fusion processing method for improving angle estimation accuracy according to claim 2 or 3, characterized in that: center spacing Keep it as small as possible, so as not to affect the rotation of the radar and camera.

5. The fusion processing method for improving angle estimation accuracy according to claim 1, characterized in that: In step 2, and To obtain the target's coarse angle information, the radar sends the coarse angle to the electro-optical camera.

6. The fusion processing method for improving angle estimation accuracy according to claim 1, characterized in that: In step 3, the image center position in the image plane. The angle is Corresponding to the optical axis direction, the target center pixel Position coordinates are , for relative pixel offset, and Positive and negative basis The quadrant in the imaging plane coordinate system is determined.

7. The fusion processing method for improving angle estimation accuracy according to claim 1, characterized in that: In step 4, the imaging plane intersects the optical axis at... Point, the center of the camera lens is The camera focus is The camera focal length is Target center point The object distance is The center pixel of the target image The image distance is , arrive The distance is the image height offset vertical angle It is the angle between the optical axis and the horizontal plane.

8. The fusion processing method for improving angle estimation accuracy according to claim 7, characterized in that: In step 4, The vertical angular offset of the target; ,in, , ; , for offset in the vertical direction distance, This refers to the pixel size in the vertical direction.

9. The fusion processing method for improving angle estimation accuracy according to claim 8, characterized in that: In step 4, the target horizontal angle is offset. ,in, , ; , for Horizontal offset distance, This refers to the pixel size in the horizontal direction.

10. The fusion processing method for improving angle estimation accuracy according to claim 1, characterized in that: In step 5, based on the precise angle and radar measurement distance This allows us to obtain the target's precise location information.

Images