A millimeter wave radar power line detection and three-dimensional reconstruction method
By using millimeter-wave radar for power line detection and 3D reconstruction, the problem of automatic detection and 3D display of power lines in helicopter collision avoidance systems has been solved, enabling efficient, automated, and visualized detection under adverse weather conditions.
Patent Information
- Application Number
- CN202210338295.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-04-01
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2042-04-01
AI Technical Summary
In existing technologies, helicopter collision avoidance systems face challenges in automatically detecting and displaying targets such as power lines in three dimensions. The level of automation and visualization is low, and the detection effect is particularly poor under adverse weather conditions.
Millimeter-wave radar is used for power line detection and 3D reconstruction. Through beam sharpening, point and surface target detection, Hough transform and rationality verification, combined with coordinate transformation and data alignment, the automatic identification and 3D rendering of power line features are achieved.
It enables efficient automatic detection and 3D visualization of power lines under adverse weather conditions, improving the driver's user experience.
Smart Images

Figure CN114814847B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of information perception and recognition, a detection and recognition method of global situational awareness in the application of millimeter-wave radar to detect power lines and a three-dimensional reconstruction method of radar data, which is suitable for the application of helicopter collision avoidance and unmanned aerial vehicle collision avoidance systems. Background Art
[0002] At present, power lines, iron towers, tall buildings, mountains, jungles, etc. pose a great threat to helicopter flight safety. The demand for helicopter collision avoidance technology is becoming more and more urgent. Optical equipment and laser radar are widely used due to their low prices, but they have poor environmental adaptability and are greatly affected by weather conditions. The probability of detection decreases under meteorological conditions such as smoke, fog, haze, rain, and snow, affecting normal detection; in addition, there are problems of difficulty in detecting targets such as power lines. Millimeter-wave radar has obvious advantages in power line detection and is not affected by weather and meteorology; however, there are still certain difficulties in automatic detection and three-dimensional display of power lines, and the degree of automation and visualization is low, which limits the driver's experience. Summary of the invention
[0003] In order to solve the problems in the background technology, the present invention proposes a millimeter wave radar power line detection and three-dimensional reconstruction method, which involves millimeter wave radar detection, beam sharpening processing of radar data, point and surface target detection and extraction, power line feature detection, Hough transform and rationality test, coordinate conversion and data alignment, three-dimensional reconstruction and three-dimensional rendering, etc.
[0004] The technical solution adopted by the present invention is:
[0005] A millimeter wave radar power line detection and three-dimensional reconstruction method, comprising the following steps:
[0006] (1) A millimeter-wave radar is used to transmit a linear frequency modulation waveform electromagnetic wave. After receiving the millimeter-wave radar, pulse compression demodulation is performed to obtain one-dimensional distance energy image data. The radar beam azimuth and elevation scanning is then used to synthesize the azimuth-elevation-range-energy three-dimensional radar image data of radar real aperture imaging;
[0007] (2) For the three-dimensional radar image data, a convolution inversion algorithm based on the radar antenna pattern is used to perform beam sharpening processing to obtain high-resolution radar image data;
[0008] (3) Setting thresholds to extract strong scattering point obstacle targets and large-area reflection area targets from high-resolution radar image data;
[0009] (4) Among the strong scattering point obstacle targets, the power line characteristic targets are selected according to the Bragg scattering distribution characteristics on the millimeter wave power lines;
[0010] (5) Performing Hough transform on the target image of strong scattering points that meet the characteristics of power lines, obtaining multiple target sets distributed in straight line segments, including power lines and lighthouses. For multiple target sets, combining target pitch information, distance information, angle information, energy information, and the number of scattering points in the straight line segment, making a rationality judgment on the straight line segment that meets the characteristics of the power line, and extracting the power line target;
[0011] (6) For high-resolution radar image data, strong scattering point obstacle targets, large-area reflection area targets and power line targets, radar coordinates-helicopter geodetic coordinates-geocentric rectangular coordinates are converted to obtain radar three-dimensional data in the geocentric coordinate system; and the radar grid data of radar range-azimuth-altitude in the geocentric coordinate system is obtained by using aggregation and interpolation algorithms;
[0012] (7) Perform three-dimensional reconstruction based on feature point growth on the radar grid data to obtain a three-dimensional scene data model; and superimpose obstacle target information, large-area target information, and power line information for three-dimensional rendering to achieve three-dimensional display of feature targets.
[0013] Wherein, the step (2) specifically comprises the following steps:
[0014] (201) The truncated singular value decomposition technique is used to preprocess the three-dimensional radar image data. The result of the truncated singular value method can be expressed as:
[0015] H - y=H - Hx+H - n
[0016] Where H represents the antenna pattern, y represents the antenna echo, x represents the original distribution of the target scattering coefficient, n represents the noise, and H - represents the inverse matrix of the antenna pattern H;
[0017] (202) The generalized cross validation function GCV is used to select the truncation parameter, and k is determined by the minimum point of the GCV function. The problem described is:
[0018]
[0019] In the formula, x k is the solution when the cutoff parameter is k, I is the identity matrix, trace(·) represents the trace of the matrix, and is the sum of the matrix diagonals, H k represents the antenna pattern at the minimum point k of the GCV function, Indicates H k The inverse matrix of
[0020] (203) Reconstruct the signal model when the truncation parameter is k and recalculate the antenna echo The reconstructed model is:
[0021]
[0022] in,
[0023] (204) For strong point targets, regularization theory is used based on the recalculated antenna echoes, target prior information is added, and an objective function with L1 norm as constraint is constructed;
[0024] (205) Taking the L1 norm as the regularization constraint and the minimum error energy as the fidelity constraint, the objective function is optimized and the related inverse problem is transformed into the following optimization problem:
[0025]
[0026] in, Indicates the minimum value of x when the right side of the equation takes the minimum value, argmin x represents the value of variable x that minimizes the objective function, and λ is a constant variable;
[0027] (206) Finally, the iterative weighted least squares method is used to solve the objective function; the optimal solution is:
[0028]
[0029] In the formula, represents the transpose of H, W represents the weight factor, W = diag|f(x)|, diag represents a diagonal matrix, where
[0030]
[0031] Where ξ→0 is a number tending to 0.
[0032] Wherein, the step (5) specifically comprises the following steps:
[0033] (501) coordinate transformation is performed on the image of the strong scattering point obstacle target that meets the characteristics of the power line, and the slant distance of each azimuth is converted into a vertical distance. After processing, the linear target appears as a straight line segment in the image;
[0034] (502) dividing the image processed in step (503) into blocks according to the distance direction and taking thresholds, setting different thresholds for different distance blocks, and the target is the one whose distance is greater than the threshold in each distance block;
[0035] (503) For the target of each distance block after screening in step (502), a single point and all points on a straight line are condensed into one point using Hough transform, and the formula is:
[0036] ρ=x cosθ+y sinθ
[0037] Among them, (x, y) represents the horizontal and vertical coordinates of a point in the rectangular coordinate system, θ represents the angle between the point (x, y) and the origin of the coordinate system, and ρ represents the distance between the condensed point and the origin of the coordinate system; in the parameter space, a point is mapped to a cosine curve, and a straight line is mapped to a cluster of cosine curves passing through the same point;
[0038] (504) By detecting the maximum value of the parameter space and mapping the parameter space back to the rectangular coordinate system, the longest straight line in each distance block in the rectangular coordinate system is obtained;
[0039] (505) Select the electric lines from the straight lines obtained in step (504) according to the following principle:
[0040] 1) Maximum spacing of points on a straight line: take the maximum spacing d max , if the distance between adjacent points on the straight line is greater than d max If the straight line is not a power line,
[0041] 2) The number of Bragg scattering peaks that can be seen on the power line: if there is a Bragg scattering peak, it means it is a power line, and if there is no Bragg scattering peak, it is not a power line;
[0042] 3) Minimum spacing on a straight line: The minimum spacing criterion is formulated based on whether there is a Bragg scattering peak on the straight line and the position of the Bragg scattering peak. If the spacing between adjacent points on the straight line is less than the minimum spacing, it is not an electric line;
[0043] If the above three conditions are met at the same time, the power line is selected.
[0044] The specific determination process of the number of Bragg scattering peaks that can be seen on the power line is as follows:
[0045] Use the position of the Bragg scattering peak to find the corresponding point on the straight line, that is, the positions of the three Bragg peaks of the power line echo are -14°, 0°, and 14°; use the energy difference between the Bragg secondary peak and the main peak to further determine, that is, the secondary peak is 7dB lower than the main peak; in plain terrain, the relationship between the distance between the power line and the radar and the length of the power line is:
[0046] The maximum distance that radar can detect power lines is R max ≈2L, L is the length of the electric line, and the three Braggs are at the same pitch angle:
[0047] when When , three Bragg scattering peaks appear on the power line, where R represents the distance between the power line and the radar;
[0048] When θ is When , three Bragg scattering peaks appear, where θ represents the elevation angle between the electric line and the radar;
[0049] When θ is When , two Bragg scattering peaks appear;
[0050] When θ is A Bragg scattering peak appears;
[0051] When θ is greater than There is no Bragg scattering peak.
[0052] The advantages of the present invention compared to the background technology are:
[0053] The present invention adopts millimeter wave real aperture imaging technology, which has strong anti-interference and good environmental adaptability;
[0054] The present invention can automatically identify obstacle targets such as power lines through feature point enhancement and power line detection algorithms;
[0055] The present invention adopts three-dimensional reconstruction of radar data to enhance the visualization degree of the target. BRIEF DESCRIPTION OF THE DRAWINGS
[0056] Figure 1 It is a flow chart of the present invention.
[0057] Figure 2 This is a flow chart of the beam sharpening algorithm of the present invention. DETAILED DESCRIPTION
[0058] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments.
[0059] A millimeter wave radar power line detection and three-dimensional reconstruction method, comprising the following steps:
[0060] (1) A millimeter-wave radar is used to transmit a linear frequency modulation waveform electromagnetic wave. After the radar receives it, pulse compression demodulation is performed to obtain one-dimensional distance energy image data; through the azimuth and elevation scanning of the radar beam, the azimuth-elevation-distance-energy three-dimensional radar image data of radar real aperture imaging is synthesized. The radar performs classic algorithms such as attitude correction and motion compensation on the three-dimensional data to eliminate the influence of airborne motion conditions and continuously perform three-dimensional imaging; and the three-dimensional image data is superimposed and updated according to the platform motion trajectory.
[0061] (2) For the three-dimensional radar image data, a convolution inversion algorithm based on the radar antenna pattern is used to perform beam sharpening processing to obtain high-resolution radar image data;
[0062] The model of the signal from input to output is written in convolution form
[0063] y(t)=x(t)*a(t)+n(t)
[0064] The expression written in the frequency domain after Fourier transform is
[0065] Y(ω)=X(ω)A(ω)+N(ω)
[0066] Written in the form of deconvolution solution:
[0067]
[0068] If [h -m , ..., h0, ..., h m ] represents the sampling of the antenna pattern, then the convolution matrix can be expressed as:
[0069]
[0070] Among them, considering that the antenna beam is scanned in and out, it will be affected by the half beam outside the scene. When constructing the convolution matrix, the number of columns is made to be one more than the number of scene points by the number of points of the antenna beam.
[0071] When the echo matrix and antenna measurement matrix are known, the problem of improving the azimuth resolution is transformed into a corresponding inversion process. In theory, the H inverse can be used to achieve target recovery and improve the azimuth resolution. The methods for improving the resolution are as follows: Figure 2 shown.
[0072] The radar azimuth echo data is in the form of the convolution of the antenna pattern and the target scattering coefficient. The convolution inversion process is the process of realizing the target reconstruction convolution model.
[0073] (201) The truncated singular value decomposition (TSVD) technique is used to preprocess the three-dimensional radar image data to suppress the noise amplification in the inversion process and reduce the interference of noise on the target information. The result of the truncated singular value method can be expressed as:
[0074] H - y=H - Hx+H - n
[0075] Where H represents the antenna pattern, y represents the antenna echo, x represents the original distribution of the target scattering coefficient, n represents the noise, and H - represents the inverse matrix of the antenna pattern H;
[0076] (202) The generalized cross validation (GCV) function is used to select the truncation parameter, and k is determined by the minimum point of the GCV function. The problem described is:
[0077]
[0078] Where x k is the solution when the cutoff parameter is k, I is the identity matrix, trace(·) represents the trace of the matrix, and is the sum of the matrix diagonals, H k represents the antenna pattern at the minimum point k of the GCV function, H k - Indicates H k The inverse matrix of
[0079] (203) Reconstruct the signal model when the truncation parameter is k and recalculate the antenna echo The reconstructed model is:
[0080]
[0081] in,
[0082] (204) For strong point targets, regularization theory is used based on the recalculated antenna echoes, target prior information is added, and an objective function with L1 norm as constraint is constructed;
[0083] (205) Taking the L1 norm as the regularization constraint and the minimum error energy as the fidelity constraint, the objective function is optimized and the related inverse problem is transformed into the following optimization problem:
[0084]
[0085] in, Indicates the minimum value of x when the right side of the equation takes the minimum value, argmin x represents the value of variable x that minimizes the objective function, and λ is a constant variable;
[0086] (206) Finally, the iterative weighted least squares method is used to solve the objective function; therefore, the optimal solution is:
[0087]
[0088] In the formula, express The transpose of , W represents the weight factor, W = diag | f (x) |, diag represents the diagonal matrix, where,
[0089]
[0090] Where ξ→0 is a number tending to 0.
[0091] (207) Since the weight factor is related to the target scattering coefficient itself, the weight factor needs to be updated every time a calculation result is obtained, and the initial value needs to be defined. The forward-looking super-resolution imaging in a low signal-to-noise ratio environment is achieved through the "remove first and add later" strategy.
[0092] (3) Setting thresholds to extract strong scattering point obstacle targets and large-area reflection area targets from high-resolution radar image data;
[0093] (4) Among the strong scattering point obstacle targets, the power line characteristic targets are selected according to the Bragg scattering distribution characteristics on the millimeter wave power lines;
[0094] (5) Performing Hough transform on the target image of strong scattering points that meet the characteristics of power lines, obtaining multiple target sets distributed in straight line segments, including power lines and lighthouses. For multiple target sets, combining target pitch information, distance information, angle information, energy information, and the number of scattering points in the straight line segment, making a rationality judgment on the straight line segment that meets the characteristics of the power line, and extracting the power line target;
[0095] After coordinate transformation and preprocessing, the radar image of the power line appears as straight line segments in the image. These straight line segments are contained in the image fragments. At this time, the Hough transform can be directly used to detect these straight line segments. Hough transform detection of straight lines requires that the pixels are on a straight line, but there is no requirement for the line width and the density of the points on the line. After appropriate preprocessing, the power lines can be easily detected, but at the same time, the straight line segments composed of noise points exceeding the threshold will also be detected. Then it is necessary to classify the detected straight line segments, determine which are power lines, and pick these line segments out. In order not to lose the power line information, as many straight line segments as possible should be detected, but doing so will cause the amount of calculation to rise sharply, which is not conducive to subsequent classification processing. Considering that the detected points are peak points within the local range of the image, only these peak points can be taken during the detection process. After multiple simulation experiments, it is determined that 2% to 5% of the peak points in each image are taken to take into account the detection probability and the amount of calculation.
[0096] The specific steps include:
[0097] (501) coordinate transformation is performed on the image of the strong scattering point obstacle target that meets the characteristics of the power line, and the slant distance of each azimuth is converted into a vertical distance. After processing, the linear target appears as a straight line segment in the image;
[0098] (502) dividing the image processed in step (503) into blocks according to the distance direction and taking thresholds, setting different thresholds for different distance blocks, and the target is the one whose distance is greater than the threshold in each distance block;
[0099] (503) For the target of each distance block after screening in step (502), a single point and all points on a straight line are condensed into one point using Hough transform, and the formula is:
[0100] ρ=x cosθ+y sinθ
[0101] Among them, (x, y) represents the horizontal and vertical coordinates of a point in the rectangular coordinate system, θ represents the angle between the point (x, y) and the origin of the coordinate system, and ρ represents the distance between the condensed point and the origin of the coordinate system; in the parameter space, a point is mapped to a cosine curve, and a straight line is mapped to a cluster of cosine curves passing through the same point;
[0102] (504) By detecting the maximum value of the parameter space and mapping the parameter space back to the rectangular coordinate system, the longest straight line in each distance block in the rectangular coordinate system is obtained;
[0103] (505) Select the electric lines from the straight lines obtained in step (504) according to the following principle:
[0104] 1) Maximum spacing of points on a straight line: take the maximum spacing d max , if the distance between adjacent points on the straight line is greater than d max If the straight line is not a power line,
[0105] 2) The number of Bragg scattering peaks that can be seen on the power line:
[0106] Use the position of the Bragg scattering peak to find the corresponding point on the straight line, that is, the positions of the three Bragg peaks of the power line echo are -14°, 0°, and 14°; use the energy difference between the Bragg secondary peak and the main peak to further determine, that is, the secondary peak is 7dB lower than the main peak; in plain terrain, the relationship between the distance between the power line and the radar and the length of the power line is:
[0107] The maximum distance that radar can detect power lines is R max ≈2L, L is the length of the electric line, and the three Braggs are at the same pitch angle:
[0108] when When , three Bragg scattering peaks appear on the power line, where R represents the distance between the power line and the radar;
[0109] When θ is When , three Bragg scattering peaks appear, where θ represents the elevation angle between the electric line and the radar;
[0110] When θ is When , two Bragg scattering peaks appear;
[0111] When θ is A Bragg scattering peak appears;
[0112] When θ is greater than When , there is no Bragg scattering peak;
[0113] If Bragg scattering peaks appear, it means that there are electric lines, and if there are no Bragg scattering peaks, it is not an electric line;
[0114] 3) Minimum spacing on a straight line: The minimum spacing criterion is formulated based on whether there is a Bragg scattering peak on the straight line and the position of the Bragg scattering peak. If the spacing between adjacent points on the straight line is less than the minimum spacing, it is not an electric line;
[0115] If the above three conditions are met at the same time, the power line is selected.
[0116] (6) For high-resolution radar image data, strong scattering point obstacle targets, large-area reflection area targets and power line targets, radar coordinates-helicopter geodetic coordinates-geocentric rectangular coordinates are converted to obtain radar three-dimensional data in the geocentric coordinate system; the coordinate conversion process includes three-dimensional scanning data of the radar polar coordinate system -> establishment of the radar local rectangular coordinate system -> Cartesian product space rectangular coordinates -> rotation and translation -> WGS84 geodetic coordinates -> Cesium three-dimensional rendering, and output of geocentric coordinate system radar three-dimensional data; and the aggregation and interpolation algorithms are used to realize the radar grid data of radar range-azimuth-altitude in the geocentric coordinate system.
[0117] (7) Perform 3D reconstruction based on feature point growth for radar grid data to obtain a 3D scene data model; superimpose obstacle target information, large area target information and power line information for 3D rendering to achieve 3D display of feature targets. Cesium 3D rendering is a JavaScript open source product for 3D earth and maps. It supports various geometric bodies: points, lines, surfaces, cubes, ellipsoids, cylindrical corridors, pipe diameters, and walls. It supports mouse and touch operations for zooming, rendering, inertial translation, flight, arbitrary viewing angles, and terrain collision detection.
[0118] After completing the above steps, the power line detection and three-dimensional reconstruction display modeling are completed.
Claims
1. A millimeter wave radar power line detection and three-dimensional reconstruction method, characterized in that: The following steps are involved: (1) A millimeter-wave radar is used to transmit a linear frequency modulation waveform electromagnetic wave. After receiving the millimeter-wave radar, pulse compression demodulation is performed to obtain one-dimensional distance energy image data. The radar beam azimuth and elevation scanning is then used to synthesize the azimuth-elevation-range-energy three-dimensional radar image data of radar real aperture imaging; (2) For the three-dimensional radar image data, a convolution inversion algorithm based on the radar antenna pattern is used to perform beam sharpening processing to obtain high-resolution radar image data; (3) Setting thresholds to extract strong scattering point obstacle targets and large-area reflection area targets from high-resolution radar image data; (4) Among the strong scattering point obstacle targets, the power line characteristic targets are selected according to the Bragg scattering distribution characteristics on the millimeter wave power lines; (5) Performing Hough transform on the target image of strong scattering points that meet the characteristics of power lines, obtaining multiple target sets distributed in straight line segments, including power lines and lighthouses. For multiple target sets, combining target pitch information, distance information, angle information, energy information, and the number of scattering points in the straight line segment, making a rationality judgment on the straight line segment that meets the characteristics of the power line, and extracting the power line target; (6) For high-resolution radar image data, strong scattering point obstacle targets, large-area reflection area targets and power line targets, radar coordinates-helicopter geodetic coordinates-geocentric rectangular coordinates are converted to obtain radar three-dimensional data in the geocentric coordinate system; and the radar grid data of radar range-azimuth-altitude in the geocentric coordinate system is obtained by using aggregation and interpolation algorithms; (7) Perform three-dimensional reconstruction based on feature point growth on radar grid data to obtain a three-dimensional scene data model; and superimpose obstacle target information, large area target information, and power line information for three-dimensional rendering to achieve three-dimensional display of feature targets; Wherein, the step (2) specifically comprises the following steps: (201) The truncated singular value decomposition technique is used to preprocess the three-dimensional radar image data. The result of the truncated singular value method can be expressed as: H - y=H - Hx+H - n Where H represents the antenna pattern, y represents the antenna echo, x represents the original distribution of the target scattering coefficient, n represents the noise, and H - represents the inverse matrix of the antenna pattern H; (202) The generalized cross validation function GCV is used to select the truncation parameter, and k is determined by the minimum point of the GCV function. The problem described is: In the formula, x k is the solution when the cutoff parameter is k, I is the identity matrix, trace(·) represents the trace of the matrix, and is the sum of the matrix diagonals, H k represents the antenna pattern at the minimum point k of the GCV function, Indicates H k The inverse matrix of (203) Reconstruct the signal model when the truncation parameter is k and recalculate the antenna echo The reconstructed model is: in, (204) For strong point targets, regularization theory is used based on the recalculated antenna echoes, target prior information is added, and an objective function with L1 norm as constraint is constructed; (205) Taking the L1 norm as the regularization constraint and the minimum error energy as the fidelity constraint, the objective function is optimized and the related inverse problem is transformed into the following optimization problem: in, Indicates the minimum value of x when the right side of the equation takes the minimum value, argmin x represents the value of variable x that minimizes the objective function, and λ is a constant variable; (206) Finally, the iterative weighted least squares method is used to solve the objective function; the optimal solution is: In the formula, express The transpose of , W represents the weight factor, W = diag | f (x) |, diag represents the diagonal matrix, where, Among them, ξ→0 is a number tending to 0; Wherein, the step (5) specifically comprises the following steps: (501) coordinate transformation is performed on the strong scattering point obstacle target image that meets the characteristics of the power line, and the slant distance of each azimuth is converted into a vertical distance. After processing, the linear target appears as a straight line segment in the image; (502) dividing the image processed in step (503) into blocks according to the distance direction and taking thresholds, setting different thresholds for different distance blocks, and the target is the one whose distance is greater than the threshold in each distance block; (503) For the target of each distance block after screening in step (502), a single point and all points on a straight line are condensed into one point using Hough transform, and the formula is: ρ=xcosθ+ysinθ Among them, (x, y) represents the horizontal and vertical coordinates of a point in the rectangular coordinate system, θ represents the angle between the point (x, y) and the origin of the coordinate system, and ρ represents the distance between the condensed point and the origin of the coordinate system; in the parameter space, a point is mapped to a cosine curve, and a straight line is mapped to a cluster of cosine curves passing through the same point; (504) By detecting the maximum value of the parameter space and mapping the parameter space back to the rectangular coordinate system, the longest straight line in each distance block in the rectangular coordinate system is obtained; (505) Select the electric lines from the straight lines obtained in step (504) according to the following principle: 1) Maximum spacing of points on a straight line: take the maximum spacing d max , if the distance between adjacent points on the straight line is greater than d max If the straight line is not a power line, 2) The number of Bragg scattering peaks that can be seen on the power line: if there is a Bragg scattering peak, it means it is a power line, and if there is no Bragg scattering peak, it is not a power line; 3) Minimum spacing on a straight line: The minimum spacing criterion is formulated based on whether there is a Bragg scattering peak on the straight line and the position of the Bragg scattering peak. If the spacing between adjacent points on the straight line is less than the minimum spacing, it is not an electric line; If the above three conditions are met at the same time, the power line is selected.
2. A millimeter wave radar power line detection and three-dimensional reconstruction method according to claim 1, characterized in that: The specific determination process of the number of Bragg scattering peaks that can be seen on the power line is as follows: Use the position of the Bragg scattering peak to find the corresponding point on the straight line, that is, the positions of the three Bragg peaks of the power line echo are -14°, 0°, and 14°; use the energy difference between the Bragg secondary peak and the main peak to further determine, that is, the secondary peak is 7dB lower than the main peak; in plain terrain, the relationship between the distance between the power line and the radar and the length of the power line is: The maximum distance that the radar can detect the power line is R max ≈2L, L is the length of the electric line, and the three Braggs are at the same pitch angle: when When , three Bragg scattering peaks appear on the power line, where R represents the distance between the power line and the radar; When θ is When , three Bragg scattering peaks appear, where θ represents the elevation angle between the electric line and the radar; When θ is When , two Bragg scattering peaks appear; When θ is A Bragg scattering peak appears; When θ is greater than There is no Bragg scattering peak.
Citation Information
Patent Citations
Electric-power line detection method and system based on millimeter wave radar decision tree classification
CN106529416A
Scanning radar pitching-azimuth two-dimensional super-resolution method
CN113064165A
System and method for 3D radar image rendering
US20080074312A1
