Ground penetrating radar shallow target detection method and device and storage medium

Abstract

The invention discloses a ground penetrating radar shallow target detection method and device and a medium, and the method comprises the steps of data collection, zero offset correction processing, median taking from data and the like. The contour line algorithm can completely and clearly describe the forms of the target and the weaker target so as to extract the complete edge curve of the target. A traditional lower opening scanning algorithm needs to distinguish lambda-shaped, X-shaped and V-shaped intersection, but the method can avoid the difficult points, and only needs to find point segments conforming to normal lower opening characteristics.

Description

Technical Field

[0001] The present invention belongs to the field of ground penetrating radar research, and particularly relates to a method, device and storage medium for detecting shallow weak targets by ground penetrating radar in this field. Background Art

[0002] When using a ground penetrating radar for target detection, if the environment is relatively complex and the target is weak, the target will be missed during background segmentation or target extraction. The traditional downward-opening scanning algorithm needs to distinguish λ-shaped, X-shaped, and V-shaped intersections, and the operation is relatively difficult. Summary of the Invention

[0003] In order to solve the technical problems that the existing algorithms for detecting hyperbolic targets lose and miss targets when the targets are weak, the present invention provides a method, device and storage medium for detecting shallow targets by ground penetrating radar, which can more accurately find hyperbolic targets.

[0004] The present invention adopts the following technical solutions:

[0005] A method for detecting shallow targets by ground penetrating radar, which is improved in that it includes the following steps:

[0006] Step 1, using a ground penetrating radar device to collect data in a test field where small targets are buried;

[0007] Step 2, first performing zero-offset correction processing on the collected original radar data;

[0008] Step 3, taking the median in the data after zero-offset correction and using the median as the height threshold for the contour line algorithm;

[0009] Step 4, using the four-point grid method to calculate the height value corresponding to each data of the four-point grid, and screening the position points that meet the height conditions through the height threshold in Step 3:

[0010]

[0011] In the above formula, level refers to the height threshold, v(k) is the k-th vertex coordinate, the k value is 1, 2, 3 or 4, and next = mod(k, 4) + 1;

[0012] x = p(k, 1) + t * (p(next, 1) - p(k, 1)

[0013] y = p(k, 2) + t * (p(next, 2) - p(k, 2)

[0014] x and y are the coordinate points of the isovalue points, p(k, 1) is the vertex coordinate at the position of k, 1, p(k, 2) is the vertex coordinate at the position of k, 2, p(next, 1) is the vertex coordinate at the position of next, 1, and p(next, 2) is the vertex coordinate at the position of next, 2;

[0015] Step 5: Connect the points at the same height, and filter out the position points where the number of connected points is less than 400;

[0016] Step 6: Set the values of the filtered position points to 1 for binarization processing;

[0017] Step 7: Traverse the binarized image, define each continuous segment of 1 in each row as a point segment, and record the starting position, ending position, and length of the point segment;

[0018] Step 8: Traverse all the point segment information, respectively find the repeated point segments a and b on the left and right of the point segment in the upper row and the lower row, or find that the starting position of the point segment in the upper row is separated by 1 from the ending position of the point segment in the lower row, and the ending position of the point segment in the upper row is separated by 1 from the starting position of the point segment in the lower row, and record the starting positions of the downward-opening point segments that meet these two conditions;

[0019] Step 9: Traverse the left and right sides of the point segments that meet the downward-opening condition respectively. Find continuous repeated point segments on the left and right sides, and if the number of continuous point segments on both sides is greater than 10, then confirm that the downward-opening hyperbola is the target.

[0020] Further, in Step 7, the number of 1s in the point segment is greater than or equal to 2.

[0021] Further, in Step 8, a and b are not a single point segment and there is an interval in the middle.

[0022] An improved ground penetrating radar shallow target detection device includes: a processor; a memory storing executable instructions of the processor; the processor is configured to execute the steps of the above method by executing the executable instructions.

[0023] An improved computer-readable storage medium for storing a program, wherein the program, when executed, implements the steps of the above method.

[0024] The beneficial effects of the present invention are:

[0025] The method disclosed by the present invention combines the contour algorithm and the lower-opening clustering algorithm. The contour algorithm can completely and clearly depict the shapes of the target and the weaker targets, so as to extract the complete edge curve of the target. The traditional lower-opening scanning algorithm needs to distinguish λ-shaped, X-shaped, and V-shaped intersections. The method of the present invention can avoid these difficulties and only needs to find the point segments that conform to the characteristics of normal lower openings. Description of the Drawings

[0026] Figure 1 is the original data map collected by the ground penetrating radar;

[0027] Figure 2 is the data map after being screened by the height threshold in step 4;

[0028] Figure 3 is the result map after lower-opening clustering in step 8. Detailed Embodiments

[0029] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0030] Embodiment 1. This embodiment discloses a method for detecting shallow targets by a ground penetrating radar. First, the contour algorithm is used to extract the complete edge curve of the target, then the data is converted into a binary image, and then the hyperbola characteristics of the target are used for screening, so as to accurately and completely extract the hyperbola target. The specific steps are as follows:

[0031] Step 1, use a ground penetrating radar device to collect data in a test field where small targets are buried; the original data collected by the ground penetrating radar is as Figure 1 shown;

[0032] Step 2, perform zero-offset correction processing on the collected original radar data;

[0033] Step 3, take the median value in the data after zero-offset correction, and use this median value as the height threshold of the contour algorithm;

[0034] Step 4, use the four-point grid method to calculate the height value corresponding to each data of the four-point grid, and screen the position points that meet the height conditions through the height threshold in step 3:

[0035]

[0036] In the above formula, level refers to the height threshold, v(k) is the coordinate of the kth vertex, the value of k is 1, 2, 3 or 4, and next = mod(k,4)+1;

[0037] x = p(k, 1) + t * (p(next, 1) - p(k, 1))

[0038] y = p(k, 2) + t * (p(next, 2) - p(k, 2))

[0039] x and y are the coordinate points of the equal - value points, p(k, 1) is the vertex coordinate at the position of k, 1, p(k, 2) is the vertex coordinate at the position of k, 2, p(next, 1) is the vertex coordinate at the position of next, 1, p(next, 2) is the vertex coordinate at the position of next, 2; the data after the height threshold screening is as Figure 2 shown;

[0040] Step 5: Connect the points at the same height to each other, and filter out the position points where the number of connected points is less than 400;

[0041] Step 6: Set the values of the filtered position points to 1 for binarization processing;

[0042] Step 7: Traverse the binarized image, define each continuous segment of 1 (the number of 1s is greater than or equal to 2) in each row as a point segment, and record the starting position, ending position, and length of the point segment;

[0043] Step 8: Traverse all the point segment information, respectively find the repeated point segments a and b (a and b are not the same point segment, there is an interval in the middle) on the left and right of the point segment in the upper row and the lower row, or find that the starting position of the point segment in the upper row and the ending position of the point segment in the lower row are separated by 1, and the ending position of the point segment in the upper row and the starting position of the point segment in the lower row are separated by 1, and record the starting positions of the downward - opening point segments that meet these two conditions; the result after downward - opening clustering is as Figure 3 shown;

[0044] Step 9: Traverse the left and right sides of the point segments that meet the downward - opening conditions respectively, find continuous repeated point segments on the left and right sides, and if the number of continuous point segments on both sides is greater than 10, then confirm that the downward - opening hyperbola is the target.

[0045] This embodiment also discloses a ground - penetrating radar shallow - target detection device, including: a processor; a memory in which executable instructions of the processor are stored; the processor is configured to execute the steps of the above - mentioned method by executing the executable instructions. And a computer - readable storage medium for storing a program, the program when executed implements the steps of the above - mentioned method.

Claims

1. A method for detecting shallow targets by ground penetrating radar, characterized in that, It includes the following steps: Step 1: Use a ground penetrating radar device to collect data at a test site where small targets are buried; Step 2: First perform zero-offset correction processing on the collected raw radar data; Step 3: Take the median value from the data after zero-offset correction and use this median value as the height threshold for the contour algorithm; Step 4: Use the four-point grid method to calculate the height value corresponding to each data point in the four-point grid, and screen out the position points that meet the height conditions through the height threshold in Step 3: In the above formula, level refers to the height threshold, v(k) is the coordinate of the k-th vertex, the k value is 1, 2, 3, or 4, and next = mod(k, 4)+1; x = p(k, 1)+t*(p(next, 1)-p(k, 1)) y = p(k, 2)+t*(p(next, 2)-p(k, 2)) x and y are the coordinate points of the isovalue points, p(k, 1) is the vertex coordinate at the position of k, 1, p(k, 2) is the vertex coordinate at the position of k, 2, p(next, 1) is the vertex coordinate at the position of next, 1, and p(next, 2) is the vertex coordinate at the position of next, 2; Step 5: Connect the points at the same height, and screen out the position points where the number of connected points is less than 400; Step 6: Set the values of the screened position points to 1 for binary processing; Step 7: Traverse the binary image, define each continuous segment of 1 in each row as a point segment, and record the start position, end position, and length of the point segment; Step 8: Traverse all the point segment information, respectively find the repeated point segments a and b on the left and right of the point segment in the previous row and the next row, or find that the start position of the point segment in the previous row and the end position of the point segment in the next row have an interval of 1, and the end position of the point segment in the previous row and the start position of the point segment in the next row have an interval of 1, and record the start positions of the lower-opening point segments that meet these two conditions; Step 9: Traverse the left and right sides of the point segments that meet the lower-opening conditions respectively. Find continuous repeated point segments on the left and right sides, and if the number of continuous point segments on both sides is greater than 10, then confirm that this lower-opening hyperbola is the target.

2. The ground penetrating radar shallow target detection method according to claim 1, characterized in that: In Step 7, the number of 1s in the point segment is greater than or equal to 2.

3. The ground penetrating radar shallow target detection method according to claim 1, characterized in that: In Step 8, a and b are not a single point segment and there is an interval in the middle.

4. A ground penetrating radar shallow target detection device, characterized in that, It includes: A processor; A memory that stores executable instructions of the processor; the processor is configured to execute the steps of the method according to Claim 1 by executing the executable instructions.

5. A computer-readable storage medium for storing a program, characterized in that, When the program is executed, it implements the steps of the method according to Claim 1.

Images