Method and device for determining underground target, equipment and storage medium

Abstract

Embodiments of the present application disclose a kind of underground target determination method and device, equipment and storage medium, method includes: obtaining the B-scan image data that underground detection acquisition is carried out by detection device;Image data is carried out column clustering separation processing, obtain multiple independent first hyperbola;According to the soil equivalent dielectric constant corresponding to each first hyperbola and Hough transform, each first hyperbola is reconstructed, to determine the second hyperbola of each first hyperbola;According to the second hyperbola and back projection algorithm, determine target focus imaging result, obtain the final positioning information of underground target.Through the above-mentioned mode, different hyperbola can be separated, each hyperbola is processed specifically, hyperbola is reconstructed by Hough transform and different soil performance, reduce clutter noise, improve underground pipeline detection accuracy, it is beneficial to accurately positioning the position information of closely arranged underground pipeline in non-uniform soil environment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radar detection, and in particular to a method and device for determining underground targets, equipment and a storage medium. BACKGROUND

[0002] Power cables have gradually replaced overhead transmission lines to become the main transmission method in urban power grid systems. However, with the continuous advancement of old city reconstruction work in China, a large number of problems of loss of laying information of old lines have occurred. The existing detection means still cannot perform visual non-destructive testing on the lines, resulting in a large number of repeated constructions in the policy promotion process, greatly increasing the difficulty of line reconstruction work, and seriously hindering the rapid development of modern cities.

[0003] In recent years, the non-destructive testing technology of underground cables based on ground penetrating radar method has attracted widespread attention. This method uses electromagnetic (EM) field to detect the change of the relative permittivity of the medium, and has been widely used in the estimation of underground target structure parameters, such as target diameter, distribution direction and target number. Traditional target detection methods usually take the measurement or estimated permittivity of a specific location as a constant of the soil in the entire survey site, so that the electromagnetic signal has the same wave speed in the entire detection area. Some widely used commercial ground penetrating radar systems, such as IDS Geo Radar and GSSI system, require the user to input a constant medium permittivity value. However, under actual working conditions, especially in tropical regions with high humidity, the non-homogeneity of the soil itself causes significant differences in electromagnetic wave speed in different parts of the soil, thereby seriously affecting the positioning accuracy of underground targets in the later stage. Gonzalez-Huici and G. Gennarelli et al. studied the variation of permittivity in different directions in soil structure, in which researchers introduced an equivalent permittivity function and preliminarily obtained the permittivity distribution rule of tropical soil through a large number of analytical calculations. On this basis, A. Bitri et al. further completed the two-dimensional ground penetrating radar data preprocessing in a horizontal layered non-homogeneous medium, and realized the two-dimensional imaging positioning of two pipelines. However, the above method can only accurately position the sparsely distributed targets in the underground space, and as the distance between targets gradually decreases, the accuracy of the traditional method also decreases sharply. According to the research results of N. R. Peplinski et al., it can be found that due to the differences in water content, composition and density, the relative permittivity of the soil at different positions in the same region may change dramatically, and the use of mismatched permittivity will lead to incorrect reconstruction of the underground space. Therefore, using the same relative permittivity to process ground penetrating radar data using the traditional method may lead to inaccurate estimation of the horizontal and vertical positions of the target. In addition, the heterogeneity of the soil will cause interference, thereby masking the reflection signals of closely arranged underground pipelines.

[0004] Therefore, there is an urgent need for an effective and accurate method to improve the detection accuracy of underground pipelines, minimize clutter noise, and accurately locate the position information of closely arranged underground pipelines in a non-uniform soil environment. SUMMARY

[0005] The main purpose of the present application is to provide a method and device for determining underground targets, equipment and storage medium, which can solve the problem of low detection accuracy of underground pipelines in the prior art.

[0006] To achieve the above-mentioned purpose, the first aspect of the present application provides a method for determining underground targets, which comprises:

[0007] Obtaining image data collected by the detection device during underground detection, wherein the image data is B-scan image data;

[0008] Performing column clustering separation processing on the image data to obtain a plurality of independent first hyperbolas, wherein the hyperbolas are used to reflect the underground position information of the underground targets;

[0009] Reconstructing each first hyperbola according to the Hough transform and the soil equivalent dielectric constant corresponding to each first hyperbola to determine a second hyperbola of each first hyperbola;

[0010] Determining a target focused imaging result according to the second hyperbola and a back projection algorithm, wherein the target focused imaging result is used to reflect the final positioning information of the underground targets.

[0011] In a feasible implementation manner, the column clustering separation processing on the image data to obtain a plurality of independent first hyperbolas comprises:

[0012] Determining the block information of each column in the image data, wherein the block is composed of a group of adjacent pixel points in the row, and the block information includes the pixel value of the pixel point of the block, the pixel coordinate of the pixel point, and the adjacent information of the block;

[0013] When the pixel value of the pixel point of the i-th column block n is greater than or equal to the preset pixel value threshold, and there is no block m adjacent to the i-th column block n in the i-1-th column, it is determined that the clustering result of the i-th column block n is the curve range of the first hyperbola, and the starting position of the first hyperbola;

[0014] When the pixel value of the pixel point of the i-th column block n is greater than or equal to the preset pixel value threshold, and there is a block m adjacent to the i-th column block n in the i-1-th column, it is determined that the clustering result of the i-th column block n is the curve range of the first hyperbola corresponding to the i-1-th column block m to which the i-th column block n belongs;

[0015] When the pixel value of the pixel point of the i-th column block n is greater than or equal to a preset pixel value threshold, and there is a block m adjacent to the i-th column block n in the i-1-th column, and there is no block w adjacent to the i-th column block in the i+1-th column, it is determined that the clustering result of the i-th column block n is that the i-th column block n belongs to the curve range of the first hyperbola corresponding to the block m in the i-1-th column, and is the termination position of the first hyperbola corresponding to the block m in the i-1-th column;

[0016] When the pixel value of the pixel point of the i-th column block n is less than the preset pixel value threshold, it is determined that the clustering result of the i-th column block n is a noise region outside the curve range of the first hyperbola.

[0017] The N blocks in the I column of the image data are scanned column by column, and according to the clustering results of the N blocks and the pixel coordinates, a plurality of independent first hyperbolas are obtained.

[0018] In a feasible implementation manner, the reconstruction of each first hyperbola according to the Hough transform and the soil equivalent dielectric constant corresponding to each first hyperbola, and the determination of a second hyperbola of each first hyperbola, comprises:

[0019] For each first hyperbola:

[0020] Randomly selecting three column curve data in the curve range of the first hyperbola;

[0021] Taking the curve coordinates corresponding to the maximum pixel value in each column curve data as reference coordinates, three groups of reference coordinates are obtained, the reference coordinates comprising the two-way propagation time of the electromagnetic wave signal and the horizontal position along the scanning track;

[0022] Substituting the three groups of reference coordinates into the mathematical expression of the Hough transform algorithm respectively, a group of hyperbolic parameters corresponding to the second hyperbola is determined, the group of hyperbolic parameters comprising at least parameters related to the depth of the underground target, the horizontal position of the underground target and the soil equivalent dielectric constant;

[0023] Determining the average value of each hyperbolic parameter in the group of hyperbolic parameters;

[0024] When the change rate of any average value is greater than or equal to a preset change rate threshold, the step of randomly selecting three column curve data in the curve range of the first hyperbola is returned to be executed until the change rate of any average value is less than the preset change rate threshold, and then the average value of each hyperbolic parameter is used to reconstruct each first hyperbola to determine the second hyperbola of the first hyperbola.

[0025] In a feasible implementation manner, the Hough transform comprises the following mathematical expression:

[0026]

[0027] a = t0

[0028] wherein,

[0029] t is the two-way propagation time of the electromagnetic wave signal, x is the horizontal position along the scan trajectory, (x0, t0) represents the center of the hyperbola, c represents the speed of light in free space, ε r represents the equivalent dielectric constant of the soil.

[0030] In a possible implementation, the target focused imaging result includes the amplitude of the imaging point, and the acquisition device is a synthetic aperture radar. The method for determining the target focused imaging result according to the second hyperbola and the back projection algorithm includes the following steps.

[0031] Determining the time delay of each imaging point in the second hyperbola to each synthetic aperture position of the synthetic aperture radar;

[0032] Taking the position at each time delay as the center, the signal data of a preset length is cut from above and below, as the echo signal of each imaging point;

[0033] Determining the signal correlation coefficient according to the scattering response amplitude vector of the echo signal and the scattering response amplitude vector of a reference signal, the reference signal being the echo signal segment at the time delay position corresponding to the middle position of the synthetic aperture;

[0034] Performing weighted superposition calculation on the signal correlation coefficient and the scattering response amplitude vector of the echo signal to determine the amplitude of each imaging point.

[0035] In a possible implementation, the signal correlation coefficient includes the following mathematical expression:

[0036]

[0037] wherein, the Cov(i, j) function represents the covariance of the vector i and the vector j, and xk represents the scattering response amplitude vector of the echo signal at the kth antenna position; represents the scattering response amplitude vector of the reference signal, and pk is the signal correlation coefficient of the scattering response amplitude vector of the echo signal at the kth antenna position and the scattering response amplitude vector of the reference signal.

[0038] In a possible implementation, the amplitude includes the following mathematical expression:

[0039]

[0040] wherein, E is the amplitude, and N Pis the number of synthetic aperture positions, k is the antenna number; p k is the signal correlation coefficient of the scattering response amplitude vector of the echo signal at the kth antenna position and the scattering response amplitude vector of the reference signal; x k is the scattering response amplitude vector of the echo signal at the kth antenna position, and S is the signal data cut up and down around the position at each time delay with a preset length.

[0041] To achieve the above object, the second aspect of the present application provides a determination device of an underground target, which comprises:

[0042] A data acquisition module is configured to acquire image data collected by the detection device during underground detection, wherein the image data is B-scan image data.

[0043] A curve separation module is configured to perform column clustering separation processing on the image data to obtain a plurality of independent first hyperbolas, wherein the hyperbolas are used to reflect underground position information of the underground target.

[0044] A curve reconstruction module is configured to reconstruct each first hyperbola according to a Hough transform and a soil equivalent dielectric constant corresponding to each first hyperbola to determine a second hyperbola of each first hyperbola.

[0045] A back projection module is configured to determine a target focused imaging result according to the second hyperbola and a back projection algorithm, wherein the target focused imaging result is used to reflect final positioning information of the underground target.

[0046] To achieve the above object, the third aspect of the present application provides a computer readable storage medium storing a computer program, wherein the computer program is executed by a processor to make the processor perform the steps of the first aspect and any feasible implementation manner.

[0047] To achieve the above object, the fourth aspect of the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the computer program is executed by the processor to make the processor perform the steps of the first aspect and any feasible implementation manner.

[0048] By adopting the embodiments of the present application, the following beneficial effects are achieved:

[0049] The application provides a method for determining an underground target, which comprises the following steps: acquiring image data collected by a detection device during underground detection, wherein the image data is B-scan image data; performing column clustering separation processing on the image data to obtain a plurality of independent first hyperbolas, wherein the hyperbolas are used to reflect underground position information of the underground target; reconstructing each first hyperbola according to a Hough transform and an equivalent dielectric constant of soil corresponding to each first hyperbola to determine a second hyperbola of each first hyperbola; and determining a target focused imaging result according to the second hyperbola and a back projection algorithm, wherein the target focused imaging result is used to reflect final positioning information of the underground target. In this way, different hyperbolas can be separated, each hyperbola can be processed specifically, the hyperbolas can be reconstructed according to a Hough transform and different soil performances, the clutter noise can be reduced, the underground pipeline detection accuracy can be improved, and the underground pipeline position information can be accurately positioned in a non-uniform soil environment. BRIEF DESCRIPTION OF DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0051] In the formula, the following applies:

[0052] Figure 1 FIG. 1 is a flowchart of a method for determining an underground target according to an embodiment of the present application;

[0053] Figure 2 FIG. 2 is another flowchart of a method for determining an underground target according to an embodiment of the present application;

[0054] Figure 3 FIG. 3 is a schematic diagram of column clustering separation processing of B-scan image data according to an embodiment of the present application;

[0055] FIG. 4(a) is a schematic diagram of a first hyperbola according to an embodiment of the present application;

[0056] FIG. 4(b) is a schematic diagram of randomly selecting three column indexes according to an embodiment of the present application;

[0057] FIG. 4(c) is a schematic diagram of selecting a row index corresponding to a maximum pixel value in the selected column according to an embodiment of the present application;

[0058] FIG. 4(d) is a schematic diagram of a second hyperbola according to an embodiment of the present application;

[0059] Figure 5A schematic diagram of a target focusing imaging result in an embodiment of the present application;

[0060] Figure 6 A structural block diagram of a determination device for an underground target in an embodiment of the present application;

[0061] Figure 7 A structural block diagram of a computer device in an embodiment of the present application. DETAILED DESCRIPTION

[0062] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0063] Please refer to Figure 1 , Figure 1 A flow chart of a determination method for an underground target in an embodiment of the present application, as shown in Figure 1 The method comprises the following steps:

[0064] 101, obtaining image data collected by a detection device during underground detection, wherein the image data is B-scan image data;

[0065] It should be noted that the underground target includes but is not limited to a cable buried underground. In order to realize the positioning of the cable buried underground, the present application takes the underground target as the cable buried underground for example to describe the embodiments. Further, when determining the underground target, a detection device is used to detect the ground surface, such as a radar or other device capable of emitting a detection beam. Through the characteristics of the echo data, the positioning of the underground cable can be realized. Therefore, the position information of the underground target can be obtained by obtaining the image data collected by the detection device during underground detection. For example, the detection device is a radar, and the image data is B-scan image data. The B-scan image data comprises a plurality of hyperbolas, and each hyperbola corresponds to an underground target.

[0066] 102, performing column clustering separation processing on the image data to obtain a plurality of independent first hyperbolas, wherein the hyperbolas are used to reflect the underground position information of the underground target;

[0067] Further, there can be more than one underground cable, and after obtaining the B-scan image data, the image data needs to be separated to obtain the hyperbola corresponding to each underground target. Therefore, the image data is separated by column clustering to obtain multiple independent hyperbolas reflecting the underground position information of the underground targets. The column clustering separation is performed by scanning the B-scan image data column by column, clustering the image data, and separating the hyperbolas by the clustering results. The first hyperbola can refer to FIG. 4(a), and the column clustering separation can obtain multiple independent hyperbolas as shown in FIG. 4(a).

[0068] 103. reconstructing each of the first hyperbolas by the Hough transform and the soil equivalent dielectric constant corresponding to each of the first hyperbolas to determine a second hyperbola of each of the first hyperbolas;

[0069] It should be noted that the dielectric properties of the soil vary significantly in different regions, and the equivalent dielectric constant of each target is also different. Therefore, the non-homogeneity of the soil needs to be further considered to significantly improve the positioning accuracy of the direct buried cable. Therefore, the characteristics of each hyperbola are extracted by combining the clustering algorithm, and the equivalent dielectric constant of the soil in the region of each hyperbola is estimated by the Hough transform, which is used to independently reconstruct each target. Finally, all the reconstructed targets are synthesized to obtain the optimized reconstructed target information. Specifically, to more accurately obtain the positioning information of the underground cable, the Hough transform and the soil equivalent dielectric constant corresponding to each of the first hyperbolas are used to reconstruct each of the first hyperbolas to obtain a more accurate second hyperbola of each of the first hyperbolas. The Hough transform is a three-parameter parabolic equation, which is used to estimate the equivalent dielectric constant of the soil above each target. Specifically, for the extracted first hyperbola, the improved Hough transform is used to fit the hyperbola based on the three-parameter parabolic equation, thereby estimating the relative dielectric constant of the soil and the depth and horizontal position of the object.

[0070] For example, the Hough transform includes the following mathematical expression:

[0071]

[0072] a = t0

[0073] In the formula,

[0074] t is the two-way propagation time of the electromagnetic wave signal, x is the horizontal position along the scanning track, (x0, t0) represents the center of the hyperbola, c represents the speed of light in free space, ε rrepresents the soil equivalent permittivity.

[0075] 104. According to the second hyperbola and the back projection algorithm, a target focused imaging result is determined, the target focused imaging result is used to reflect the final positioning information of the underground target.

[0076] Although the output of the hyperbolic fitting can be used to determine the center of the target, in order to more accurately realize the reconstruction of the structure characteristics of the underground target, it is necessary to introduce the back projection imaging method. Further, according to the second hyperbola and the back projection algorithm, a target focused imaging result is determined, the target focused imaging result is used to reflect the final positioning information of the underground target, which includes depth and horizontal distance, which can be referred to as Fig. 4(d).

[0077] The present application provides a kind of determination method of underground target, method includes: the image data that underground detection acquisition is collected by detection device, the image data is B-scan image data;Image data is carried out column cluster separation processing, obtains multiple independent first hyperbola, hyperbola is used to reflect the underground location information of underground target;According to Hough transform and the soil equivalent permittivity corresponding to each first hyperbola, each first hyperbola is reconstructed, to determine the second hyperbola of each first hyperbola;According to the second hyperbola and the back projection algorithm, a target focused imaging result is determined, the target focused imaging result is used to reflect the final positioning information of underground target. By the above-mentioned mode, different hyperbolas can be separated, each hyperbola is processed specifically, hyperbola is reconstructed by Hough transform and different soil performance, reduce clutter noise, improve underground pipeline detection precision, it is beneficial to accurately positioning the position information of closely arranged underground pipeline in non-uniform soil environment.

[0078] Please refer to Figure 2 , Figure 2 Another flow chart of a kind of determination method of underground target in the embodiment of the present application, as Figure 2 The method comprises the following steps:

[0079] 201, the image data that underground detection acquisition is collected by detection device, the image data is B-scan image data;

[0080] It should be noted that the content of step 201 is similar to that of step 101 shown in Figure 1 To avoid repetition, this place will not be described, and the specific content can be referred to the content of step 101 shown in the foregoing Figure 1 .

[0081] 202、determining block information of each column in the image data, the block being composed of a group of row-adjacent pixels, the block information including pixel values of pixels of the block, pixel coordinates of the pixels, and adjacent information of the block;

[0082] It should be noted that the Bscan image data can include pixel information of the Bscan image, the pixel information including positions of each pixel in the image, such as rows and columns (pixel coordinates) of the pixel in the image, and pixel values of the pixel. Furthermore, in order to realize extraction of a single hyperbola, the pixels in the pixel information can be divided according to the positional relationship of the pixels, so as to obtain image data of the same hyperbola. Specifically, block information of each column in the image data is determined, the block information including pixel values of pixels of the block, pixel coordinates of the pixels, and adjacent information of the block, the block being a group of row-adjacent pixels, and the block information being used to determine which block belongs to which hyperbola, or which block belongs to an element of the hyperbola, or which noise. Figure 3 For example, Figure 3 There are c1 to c 32 32 columns in total, wherein the shaded parts are blocks, there is one block in c2, and there are two blocks in c 11 There are two blocks in c3. The positional relationship between the blocks and the pixel values of the pixels can determine which block belongs to which hyperbola, or which block belongs to an element of the hyperbola, or which noise.

[0083] 203、when the pixel value of the pixel of the block n in the i-th column is greater than or equal to a preset pixel value threshold, and there is no block m adjacent to the block n in the i-th column in the i-1-th column, it is determined that the clustering result of the block n in the i-th column is the curve range of the first hyperbola, and is the starting position of the first hyperbola;

[0084] 204、when the pixel value of the pixel of the block n in the i-th column is greater than or equal to a preset pixel value threshold, and there is a block m adjacent to the block n in the i-th column in the i-1-th column, it is determined that the clustering result of the block n in the i-th column is that the block n in the i-th column belongs to the curve range of the first hyperbola corresponding to the block m in the i-1-th column;

[0085] 205、when the pixel value of the pixel of the block n in the i-th column is greater than or equal to a preset pixel value threshold, and there is a block m adjacent to the block n in the i-th column in the i-1-th column, and there is no block w adjacent to the block n in the i-th column in the i+1-th column, it is determined that the clustering result of the block n in the i-th column is that the block n in the i-th column belongs to the curve range of the first hyperbola corresponding to the block m in the i-1-th column, and is the termination position of the first hyperbola corresponding to the block m in the i-1-th column;

[0086] 206、when the pixel value of the pixel of the block n in the i-th column is less than a preset pixel value threshold, it is determined that the clustering result of the block n in the i-th column is a noise region outside the curve range of the first hyperbola.

[0087] 207. Scan each of the N blocks in column I of the image data column by column, and obtain multiple independent first hyperbolas based on the clustering results of the N blocks and the pixel coordinates;

[0088] It should be noted that the hyperbola separation and extraction technique based on clustering algorithms can extract hyperbola features from B-Scan, thereby analyzing the extracted hyperbolas separately to reduce interference from hyperbola intersection regions and environmental noise. Specifically, a column clustering algorithm is used to extract hyperbola information from the B-Scan. Figure 3 Taking the B-Scan shown as an example, the column clustering algorithm can extract hyperbolas 1 and 2 while ignoring noise 1 and 2.

[0089] For ease of description, a group of adjacent pixels in each column is called a "block". Blocks consisting of pixels with values ​​less than a threshold are called "noise," such as noise 1 (containing 2 pixels) and noise 2 (containing 5 pixels). Typically, the noise threshold *s* depends on the sensor's noise level, the radar center frequency, and the sampling frequency. The maximum value of *s* is related to the sampling frequency *f*. s Proportional to the center frequency f c Inversely proportional. For example, to suppress most noise, an ideal threshold s should be greater than the pixel value in most of the noise, but less than kf. s / f c (k is a constant).

[0090] This application uses a column clustering algorithm to cluster image data column by column. First, by analyzing the pixel values ​​of individual pixels, it determines which blocks are noise outside the hyperbola, thus removing noisy data. Then, by comparing the pixel values ​​of pixels in each block with a preset pixel value threshold, more noise data is removed. If the pixel value of a pixel in block n of column i is less than the preset pixel value threshold, it is determined to be a noise region in step 206. For example, Figure 3 Noise 1 and noise 2 exist in it.

[0091] Further, if the pixel value of the pixel point of the block n in the i-th column is greater than or equal to the preset pixel value threshold, it is determined that the block n is not noise, and the hyperbolic curve to which each block belongs is determined according to the positional relationship between the blocks. For details, refer to steps 203-206. The image data is scanned column by column from the first column to determine the clustering result of each block. i represents the column number being determined, and n represents the block being determined. When there is no adjacent block in the i-1-th column before the block n in the i-th column, it is determined that the block n in the i-th column is the starting point of the hyperbolic curve. When there is a block m adjacent to the block n in the i-th column in the i-1-th column, it is determined that the clustering result of the block n in the i-th column is that the block n in the i-th column belongs to the curve range of the first hyperbolic curve corresponding to the block m in the i-1-th column. It can be understood that if there are multiple adjacent blocks, the block n in the i-th column can be divided into each hyperbolic curve corresponding to the block m in the i-1-th column adjacent to the block n in the i-th column, so as to ensure the continuity of each hyperbolic curve. When there is a block m adjacent to the block n in the i-th column in the i-1-th column, and there is no block w adjacent to the block n in the i-th column in the i+1-th column, it is determined that the block n in the i-th column is the end point of the hyperbolic curve. It is determined that the clustering result of the block n in the i-th column is that the block n in the i-th column belongs to the curve range of the first hyperbolic curve corresponding to the block m in the i-1-th column, and the first hyperbolic curve corresponding to the block m in the i-1-th column is the end position.

[0092] Finally, the clustering result of each block in each column can be obtained, and a plurality of independent first hyperbolic curves can be obtained according to the clustering result of the N blocks and the pixel coordinates. The pixel coordinates can include (time delay, scanning distance), and each independent hyperbolic curve can be obtained by arranging the pixel coordinates of the same hyperbolic curve.

[0093] For example, Figure 3 A schematic diagram of the column clustering and separation processing of the B-scan image data in the embodiment of the present application is shown in FIG. 2. Figure 3 As shown in FIG. 2, the transceiving antenna scans from the first column c1 to the last column c 32 The adjacent blocks in the adjacent columns are classified according to the following steps:

[0094] 1) Each new hyperbolic curve region starts from the block without adjacent blocks in the previous column, that is, the block 1 in the c2 column and the block 3 in the c 15 column.

[0095] 2) When a block is adjacent to a block in any region, the block is added to the region of the adjacent block, for example, the block 2 in the c6 column needs to be added to the region to which the c5 column belongs, that is, the region to which the c2 column belongs.

[0096] 3) If a block in a column is shared by multiple regions (for example, there are two adjacent blocks in the previous column of block 4, and the two blocks belong to different regions), then the block 4 belongs to both regions to ensure the continuity of the classification process.

[0097] 4) When there is no pixel adjacent to a block in the next column, the region stops extending (for example, the block 5 in c 19 and the block 6 in c 31 ).

[0098] In addition, an additional pixel value threshold is also set, when the pixel value between a block and an existing region is less than the threshold, the region should stop expanding. In summary, after the classification is completed, the coordinates of the pixels in each region are recorded, by extracting the pixel coordinates and corresponding values in the original B-Scan, the separation and extraction of each hyperbola are realized. As shown in Figure 3 The two regions, i.e., the regions where the hyperbola 1 and the hyperbola 2 are located, are distinguished by color. Further, a single hyperbola can refer to FIG. 4(a), which is a schematic diagram of a first hyperbola in an embodiment of the present application.

[0099] 208. Reconstructing each of the first hyperbolas according to the Hough transform and the soil equivalent dielectric constant corresponding to each of the first hyperbolas to determine a second hyperbola of each of the first hyperbolas;

[0100] It should be noted that step 208 is similar to the content of step 103 shown in Figure 1 For brevity, the content of step 103 shown in Figure 1 may be referred to.

[0101] In a feasible implementation manner, step 208 can include: performing steps A1-A5 for each of the first hyperbolas:

[0102] A1. Randomly selecting three column curve data within the curve range of the first hyperbola;

[0103] A2. Taking the curve coordinates corresponding to the maximum pixel value in each column curve data as reference coordinates to obtain three groups of reference coordinates, the reference coordinates including the bidirectional propagation time of the electromagnetic wave signal and the horizontal position along the scanning track;

[0104] A3. Substituting the three groups of reference coordinates into the mathematical expression of the Hough transform algorithm respectively to determine a group of hyperbola parameters corresponding to the second hyperbola, the group of hyperbola parameters including at least parameters related to the depth of the underground target, the horizontal position of the underground target and the soil equivalent dielectric constant;

[0105] A4. Determining the average value of each hyperbola parameter in the group of hyperbola parameters;

[0106] A5. When the rate of change of any of the average values ​​is greater than or equal to a preset rate of change threshold, the process returns to the step of randomly selecting three columns of curve data within the curve range of the first hyperbola until the rate of change of any of the average values ​​is less than the preset rate of change threshold. Then, the average value of each hyperbola parameter is used to reconstruct each of the first hyperbolas to determine the second hyperbola of the first hyperbola.

[0107] It should be noted that the GPR antenna transmits electromagnetic pulses to the ground, and the receiving antenna records the time series (A-Scan) at each detection location. When the electromagnetic wave propagates to the interface between the air and soil, and between the soil and the buried cable, it reflects back a pulse signal. This is reflected in each A-Scan signal profile as a small signal peak at a distance. To ensure consistency in the calculation process, it is uniformly stipulated that the peak time of each echo signal is used as the reference time delay of the target echo, while the trough of the first echo signal is used as the zero point of the reference time.

[0108] To solve for the unknown parameters a, b, and x0 in the Hough transform and fit the first hyperbola, three columns of data are randomly selected from the first hyperbola in step A1. See Figure 4(b) for details. Figure 4(b) is a schematic diagram of randomly selecting three column indices in an embodiment of the present invention. Three different columns are randomly selected from the region matrix in Figure 4(b). Further, in step A2, the row t containing the maximum pixel value in the corresponding k-th column is selected. k This allows us to obtain the coordinates (x, y) of the point where the maximum pixel value in that column is located. k ,t k As a reference coordinate, please refer to Figure 4(c). Figure 4(c) is a schematic diagram of selecting the row index corresponding to the maximum pixel value in the selected column in an embodiment of the present invention. Since three columns are randomly selected, each column corresponds to the coordinates (x, y) of the point where the maximum pixel value is located. k ,t k This allows us to obtain the coordinates (x, y) of the points containing the three sets of maximum pixel values. k ,t k Furthermore, through step A3, the coordinates (x, y) in the three columns are... k ,t k Substituting these parameters into the Hough transform expression yields a set of hyperbolic parameters (a, b, x0). By repeating this process multiple times, multiple sets of hyperbolic parameters can be calculated. Each time the process is repeated, the average value of any one of the three hyperbolic parameters is calculated in step A4 to determine whether to stop the random selection. When the change in the average value is less than a preset rate of change threshold, such as 0.1%, the experiment is stopped, and the average value of the parameter (a, b, x0) is used in step A5. m b m x0m ) reconstructing the first hyperbola and extracting the soil equivalent permittivity in combination with the depth and horizontal position of the target. Wherein, refer to FIG. 4(d), which is a schematic diagram of a second hyperbola in an embodiment of the present application.

[0109] 209. determining a target focused imaging result according to the second hyperbola and a back projection algorithm, the target focused imaging result being used to reflect final positioning information of the underground target.

[0110] It should be noted that step 209 is similar to the content of step 104 shown in FIG. 1, and thus will not be described here again. For details, refer to the content of step 104 shown in FIG. 1. Figure 1 Figure 1 It should be noted that step 209 is similar to the content of step 104 shown in FIG. 1, and thus will not be described here again. For details, refer to the content of step 104 shown in FIG. 1.

[0111] In a feasible implementation manner, the target focused imaging result includes amplitude of an imaging point, and the acquisition device is a synthetic aperture radar. Then, step 209 can include B1-B4.

[0112] B1. determining time delay of each imaging point in the second hyperbola to each synthetic aperture position of the synthetic aperture radar;

[0113] B2. taking signal data of a preset length above and below the position as a center of each time delay as echo signal of each imaging point;

[0114] B3. determining a signal correlation coefficient according to a scattering response amplitude vector of the echo signal and a scattering response amplitude vector of a reference signal, the reference signal being an echo signal segment at a time delay position corresponding to a middle position of the synthetic aperture;

[0115] B4. performing weighted superposition calculation by using the signal correlation coefficient and the scattering response amplitude vector of the echo signal to determine amplitude of each imaging point.

[0116] It should be noted that although the output of the hyperbolic fitting can be used to determine the center of the target, in order to more accurately realize reconstruction of the structure characteristics of the underground target, a back projection imaging method (i.e. a back projection imaging method) needs to be introduced. The basic principle of the standard back projection algorithm of the ground penetrating radar is time delay-summing. However, the scattering response at each trace time delay position contains the response of all imaging points at this trace time delay position, i.e. not only the response of the to-be-imaged point, but also the interference of other imaging points. In order to solve this problem, the present application improves the traditional imaging algorithm based on the cross-correlation method, and uses the correlation coefficient of the signal at each trace delay position and the signal at the center trace delay position as the weighting coefficient of the back projection algorithm, so as to effectively suppress noise and interference.

[0117] ​Furthermore, in the cross-correlation-based back projection imaging algorithm, if the number of synthetic aperture positions is N p When imaging a specific point underground, it is necessary to first calculate the time delay from that point to each synthetic aperture position: {τ1,τ2,…,τ Np That is, step B1 determines the time delay from each imaging point in the second hyperbola to the respective synthetic aperture position of the synthetic aperture radar.

[0118] Furthermore, in step B2, a segment of effective signal can be taken above and below each time delay position as the transmitted narrow pulse segment, i.e. the part where the main energy is concentrated, as the echo signal of each imaging point; the echo signal segment at the time delay position corresponding to the middle position of the synthetic aperture is taken as the reference signal. In step B3, each signal segment is cross-correlated with the reference signal to obtain the cross-correlation signal coefficient, i.e. the signal correlation coefficient.

[0119] The length of the effective signal segment that can be taken at the top and bottom can be determined according to the following formula:

[0120]

[0121] Where L is the length of the selected signal segment, and is an odd number. S data points are taken above and below the delay position, and these, along with the data at the delay position, form the echo segment vector. The sign is... F represents the integer symbol. s f0 represents the equivalent sampling frequency and the center frequency of the transmitted signal.

[0122] The amplitude vector of the scattering response in the corresponding echo band can be expressed as:

[0123]

[0124] In the formula, x k Let s represent the scattering response amplitude vector of the echo band at the k-th antenna position. k This represents the echo vector recorded at the k-th antenna position. If an intermediate sample value exceeds the range M of the number of sampling points, or is less than 0, it is padded with 0. This step generates an L×N image about the imaging points. p A 3D matrix [x1, x2, ..., x] Np [ ] represents the reference echo vector, i.e., the scattering response value at the midpoint. If S = 0, the resulting matrix is ​​1 × N. p It is a 3D matrix, the same as the standard back projection algorithm. The echo band is the echo signal.

[0125] Furthermore, to reduce the energy of sidelobes and interference, a set of weighted coefficient vectors ρ = {ρ1, ρ2, ..., ρ3} is calculated in step B3 based on the correlation between the echo band vectors obtained at each synthetic aperture position and the reference vector.Np} wherein the calculation of the signal correlation coefficient adopts the Pearson correlation coefficient, that is, the signal correlation coefficient includes the following mathematical expression:

[0126]

[0127] wherein the Cov(i,j) function represents the covariance of the vector i and the vector j, x k represents the scattering response amplitude vector of the echo signal at the kth antenna position; represents the scattering response amplitude vector of the reference signal, and k is the signal correlation coefficient of the scattering response amplitude vector of the echo signal at the kth antenna position and the scattering response amplitude vector of the reference signal.

[0128] Finally, the coefficient weighting superposition can be performed through step B4, and the amplitude of the imaging point can be obtained, and the amplitude calculation can refer to the following mathematical expression:

[0129]

[0130] wherein E is the amplitude, N P is the number of synthetic aperture positions, and k is the antenna number; and k is the signal correlation coefficient of the scattering response amplitude vector of the echo signal at the kth antenna position and the scattering response amplitude vector of the reference signal; and k is the scattering response amplitude vector of the echo signal at the kth antenna position, and S is the signal data which is cut up and down at each time delay position with a preset length. In the above manner, the underground pipeline detection precision is effectively and accurately improved, the clutter noise is maximally reduced, the underground pipeline position information which is closely arranged is accurately positioned in a non-uniform soil environment, and finally, the target focusing imaging result can refer to Figure 5 is a schematic diagram of a target focusing imaging result in an embodiment of the present application, Figure 5 which includes the depth of the underground target and the scanning distance. By calculating the center coordinates of the focusing point, the underground cable can be accurately positioned.

[0131] A signal processing method for direct-buried cable ground penetrating radar data in a non-uniform soil environment is provided in the present application, and the two-dimensional reconstruction result is improved by considering the change of the soil dielectric constant. The key points of the present application are as follows:

[0132] (1) A multi-target hyperbolic feature separation and extraction method based on column clustering algorithm is provided, and each hyperbolic feature is processed individually to reduce the clutter caused by soil heterogeneity;

[0133] (2) Using the Hough transform technique to estimate the soil position equivalent dielectric constant of each hyperbolic curve, and combining the back projection algorithm to individually reconstruct the cable cross-sectional shape, and finally superimposing multiple focusing imaging results to complete the high-precision focusing of the underground features;

[0134] (3) Based on the cross-correlation method, the back projection algorithm is improved, and the sidelobe interference is further weakened, so as to effectively improve the detection accuracy of the underground pipeline.

[0135] The application provides a kind of underground target determination method, and the underground pipeline target is reconstructed by the improved method, and the vertical equivalent dielectric constant of the medium between antenna and direct buried cable is used to calculate electromagnetic wave velocity.Specifically, first, the two-dimensional distance profile is preprocessed by column clustering algorithm to extract hyperbolic curve.In this process, the hyperbolic curve representing different pipelines is separated and extracted into new B-Scan, while removing random clutter noise.Next, an improved Hough transform method is used to estimate the target position represented by the hyperbolic curve in each extracted B-Scan and the equivalent soil dielectric constant in the vertical direction.On this basis, the reconstruction image of each pipeline can be obtained by optimizing the dielectric constant.Finally, the above reconstruction results are fused, and the accurate position information of each pipeline target in a non-uniform soil environment can be obtained.

[0136] Please refer to Figure 6 , Figure 6 The structure block diagram of a kind of underground target determination device in the embodiment of the application is as shown in Figure 6 The device comprises:

[0137] Data acquisition module 601: for acquiring the image data collected by the underground detection device, the image data is B-scan image data;

[0138] Curve separation module 602: for column clustering separation processing to the image data, to obtain a plurality of independent first hyperbolic curves, the hyperbolic curve is used to reflect the underground position information of the underground target;

[0139] Curve reconstruction module 603: for reconstructing each first hyperbolic curve according to Hough transform and the soil equivalent dielectric constant corresponding to each first hyperbolic curve, to determine the second hyperbolic curve of each first hyperbolic curve;

[0140] Back projection module 604: for determining target focusing imaging result according to the second hyperbolic curve and back projection algorithm, the target focusing imaging result is used to reflect the final positioning information of the underground target.

[0141] It should be noted that, as Figure 6 The role of each module in the device is shown in Figure 1The steps in the method shown are similar, and to avoid repetition, they will not be elaborated here. Please refer to the relevant documentation for details. Figure 1 The content of each step in the method shown.

[0142] This invention provides a device for determining underground targets. The device includes: a data acquisition module for acquiring image data collected by a detection device during underground detection, wherein the image data is B-scan image data; a curve separation module for performing column clustering separation processing on the image data to obtain multiple independent first hyperbolas, which reflect the underground location information of the underground target; a curve reconstruction module for reconstructing each first hyperbola according to the Hough transform and the equivalent dielectric constant of the soil corresponding to each first hyperbola, thereby determining the second hyperbola for each first hyperbola; and a back projection module for determining the target focusing imaging result according to the second hyperbola and the back projection algorithm, wherein the target focusing imaging result reflects the final location information of the underground target. Through the above method, different hyperbolas can be separated, each hyperbola can be processed specifically, and the hyperbolas can be reconstructed using the Hough transform and different soil characteristics, reducing clutter noise, improving the accuracy of underground pipeline detection, and facilitating the accurate location of closely spaced underground pipelines in non-uniform soil environments.

[0143] Figure 7 An internal structural diagram of a computer device in one embodiment is shown. This computer device can specifically be a terminal or a server. Figure 7 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and may also store a computer program, which, when executed by the processor, causes the processor to perform the aforementioned methods. The internal memory may also store a computer program, which, when executed by the processor, causes the processor to perform the aforementioned methods. Those skilled in the art will understand that… Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0144] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform actions such as... Figure 1 or Figure 2 The steps of the method shown.

[0145] In one embodiment, a computer readable storage medium is provided, storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method as shown in Figure 1 or Figure 2

[0146] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware, and the program can be stored in a non-volatile computer readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments. Any reference to memory, storage, database or other medium used in the embodiments provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0147] Any combination of the technical features of the above embodiments can be made, and in order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist, it should be considered as the scope of the present application.

[0148] The above embodiments only express several embodiments of the present application, and the description is more specific and detailed, but it should not be understood as limiting the scope of the patent of the present application. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of protection of the present application. Therefore, the scope of protection of the patent of the present application should be subject to the appended claims.​

Claims

1. A method of determining an underground target, characterized in that, The method comprises: acquiring image data collected by a detection device during underground detection, the image data being B-scan image data; performing column clustering separation processing on the image data to obtain a plurality of independent first hyperbolas, the hyperbolas being used to reflect underground position information of the underground target; reconstructing each first hyperbola according to a Hough transform and a soil equivalent dielectric constant corresponding to each first hyperbola to determine a second hyperbola of each first hyperbola; determining a target focused imaging result according to the second hyperbola and a back projection algorithm, the target focused imaging result being used to reflect final positioning information of the underground target; wherein the reconstructing each first hyperbola according to a Hough transform and a soil equivalent dielectric constant corresponding to each first hyperbola to determine a second hyperbola of each first hyperbola comprises: for each first hyperbola: randomly selecting three column curve data within a curve range of the first hyperbola; taking a curve coordinate corresponding to a maximum pixel value in each column curve data as a reference coordinate to obtain three groups of reference coordinates, the reference coordinates including bidirectional propagation time of an electromagnetic wave signal and horizontal position along a scanning track; substituting the three groups of reference coordinates into a mathematical expression of the Hough transform algorithm respectively to determine a group of hyperbola parameters corresponding to the second hyperbola, the group of hyperbola parameters including at least parameters related to depth of the underground target, horizontal position of the underground target and the soil equivalent dielectric constant; determining an average value of each hyperbola parameter in the group of hyperbola parameters; when a change rate of any average value is greater than or equal to a preset change rate threshold, returning to perform the step of randomly selecting three column curve data within the curve range of the first hyperbola until the change rate of any average value is less than the preset change rate threshold, then reconstructing each first hyperbola using the average value of each hyperbola parameter to determine the second hyperbola of the first hyperbola.

2. The method of claim 1, wherein, the performing column clustering separation processing on the image data to obtain a plurality of independent first hyperbolas comprises: determining block information of each column in the image data, the block being composed of a group of adjacent pixel points in a row, the block information including pixel value of a pixel point of the block, pixel coordinate of the pixel point and adjacent information of the block; when pixel value of a pixel point of an i-th column block n is greater than or equal to a preset pixel value threshold, and there is no block m adjacent to the i-th column block n in an i-1-th column, determining that a clustering result of the i-th column block n is a curve range of a first hyperbola and is a starting position of the first hyperbola; when pixel value of a pixel point of an i-th column block n is greater than or equal to a preset pixel value threshold, and there is a block m adjacent to the i-th column block n in an i-1-th column, determining that a clustering result of the i-th column block n is that the i-th column block n belongs to a curve range of a first hyperbola corresponding to the block m in the i-1-th column. When the pixel value of the pixel point of the block n in the i-th column is greater than or equal to a preset pixel value threshold, and there is a block m adjacent to the block n in the i-th column in the (i-1)-th column, and there is no block w adjacent to the block n in the i-th column in the (i+1)-th column, it is determined that the clustering result of the block n in the i-th column is that the block n in the i-th column belongs to the curve range of the first hyperbola corresponding to the block m in the (i-1)-th column, and is the termination position of the first hyperbola corresponding to the block m in the (i-1)-th column; When the pixel value of the pixel point of the block n in the i-th column is less than the preset pixel value threshold, it is determined that the clustering result of the block n in the i-th column is a noise region outside the curve range of the first hyperbola. The N blocks in the I column of the image data are scanned column by column, and a plurality of independent first hyperbolas are obtained according to the clustering results of the N blocks and the pixel coordinates.

3. The method of claim 1, wherein, The Hough transform includes the following mathematical expression: ； In the formulae, , t is a two-way propagation time of the electromagnetic wave signal, x is a horizontal position along the scan trajectory, x 0, t 0) represents the center of the hyperbola, c represents the speed of light in free space, represents the soil equivalent permittivity.

4. The method of claim 1, wherein, The target focused imaging result includes the amplitude of the imaging point, and the detection device is a synthetic aperture radar, and the target focused imaging result is determined according to the second hyperbola and the back projection algorithm, including: Determining the time delay of each imaging point in the second hyperbola to each synthetic aperture position of the synthetic aperture radar; Taking each time delay position as the center, the signal data of a preset length is cut off upwards and downwards as the echo signal of each imaging point; According to the scattering response amplitude vector of the echo signal and the scattering response amplitude vector of the reference signal, a signal correlation coefficient is determined, and the reference signal is the echo signal segment at the time delay position corresponding to the middle position of the synthetic aperture; The signal correlation coefficient and the scattering response amplitude vector of the echo signal are used for weighted superposition calculation to determine the amplitude of each imaging point.

5. The method of claim 4, wherein, The signal correlation coefficient includes the following mathematical expression: ； wherein the Cov(i, j) function represents the covariance of vector i with vector j, x k denotes the scattering response amplitude vector of the echo signal at the k antenna position; denotes the scattering response amplitude vector of the reference signal, The amplitude includes the following mathematical expression: k is the signal correlation coefficient of the scattering response amplitude vector of the echo signal at the k antenna position and the scattering response amplitude vector of the reference signal.

6. The method of claim 4, wherein, The device includes: ； In the formula, E is the amplitude, N P is the number of synthetic aperture positions, and k is the antenna number; A data acquisition module is configured to acquire image data collected by a detection device during underground detection, wherein the image data is B-scan image data. k is the signal correlation coefficient of the scattering response amplitude vector of the echo signal at the kth k antenna position and the scattering response amplitude vector of the reference signal; x k is the scattering response amplitude vector of the echo signal at the kth k antenna position, S is the signal data of the preset length taken upward and downward from the position at each time delay.

7. A device for determining an underground target, characterized in that A curve separation module is configured to perform column clustering separation processing on the image data to obtain a plurality of independent first hyperbolas, wherein the hyperbolas are used to reflect underground position information of the underground target. A curve reconstruction module is configured to reconstruct each first hyperbola according to a Hough transform and a soil equivalent permittivity corresponding to each first hyperbola to determine a second hyperbola of each first hyperbola. A back projection module is configured to determine a target focused imaging result according to the second hyperbola and a back projection algorithm, wherein the target focused imaging result is used to reflect final positioning information of the underground target. ​ ​ The curve reconstruction module is specifically configured to: for each of the first hyperbolas: randomly selecting three columns of curve data within the curve range of the first hyperbola; taking the curve coordinates corresponding to the maximum pixel value in each column of curve data as reference coordinates to obtain three groups of reference coordinates, the reference coordinates including the two-way propagation time of the electromagnetic wave signal and the horizontal position along the scanning track; substituting the three groups of reference coordinates into the mathematical expression of the Hough transform algorithm respectively to determine a group of hyperbola parameters corresponding to the second hyperbola, the group of hyperbola parameters including at least parameters related to the depth of the underground target, the horizontal position of the underground target and the equivalent dielectric constant of the soil; determining the average value of each hyperbola parameter in the group of hyperbola parameters; when the change rate of any of the average values is greater than or equal to a preset change rate threshold, returning to the step of randomly selecting three columns of curve data within the curve range of the first hyperbola until the change rate of any of the average values is less than the preset change rate threshold, then reconstructing each of the first hyperbolas by using the average value of each hyperbola parameter to determine the second hyperbola of the first hyperbola.

8. A computer readable storage medium storing a computer program, characterized in that, The computer program is executed by the processor to enable the processor to perform the steps of the method according to any one of claims 1 to 6. 9.A computer device, comprising a memory and a processor, and characterized in that, The memory stores a computer program, and the computer program is executed by the processor to enable the processor to perform the steps of the method according to any one of claims 1 to 6.

Images