Underground cable ground penetrating radar image automatic identification method based on similarity index

Abstract

The invention relates to an underground cable ground penetrating radar image automatic identification method based on a similarity index, and the method comprises the steps: carrying out the detection of a plurality of points randomly selected in a detection region through a ground penetrating radar, and obtaining a plurality of ground penetrating radar image data; processing the ground penetrating radar image data to obtain a single-channel waveform under the electromagnetic wave radiation of the ground penetrating radar; the method comprises the following steps: detecting areas where typical media are located, collecting single-channel waveforms of the typical media, and constructing a single-channel waveform library of the typical media; calculating a similarity index between a single-channel waveform obtained by the ground penetrating radar in a region with an unknown medium type and a waveform in a single-channel waveform library of a typical medium; and determining the medium type of the point in the detection area according to the magnitude of the similarity index. According to the method, the medium type of the detection area is determined in combination with the calculation result of the similarity index, cable identification errors possibly caused by human subjective factors are reduced, meanwhile, the calculation process is simple and convenient, the required result can be rapidly obtained, and high effectiveness is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent recognition, and in particular to a method for automatically recognizing underground cable ground penetrating radar images based on a similarity index. Background Art

[0002] In recent years, with the rapid development of smart grid and urban construction, the shortage of urban land, traffic congestion, environmental pollution and other issues have become increasingly prominent. The proportion of power cables used in power transmission and distribution projects and urban distribution networks has increased significantly, and cable transmission has become one of the important ways of power transmission. Compared with overhead lines, cables have the advantages of high transmission reliability, strong anti-interference ability, and land saving.

[0003] The rapid development of power transmission and distribution projects and urban distribution networks requires confirmation of the subsurface geological conditions, underground cables and pipelines during the construction of power transmission and distribution projects. Existing large-scale power transmission and distribution line construction requires special geological surveys, but there is no corresponding survey fee for the construction of small-scale power transmission and distribution projects, and it is impossible to determine the on-site geological conditions. The on-site stratum conditions are roughly determined mainly through the collection and analysis of basic data, on-site investigations, excavation verification and other methods, which is difficult to adapt to the variability and complexity of the geological background. In the construction of power transmission and distribution projects, the use of non-destructive methods to quickly and accurately detect underground geological structures, underground cables and pipeline directions is a key issue.

[0004] Ground penetrating radar transmits high-frequency electromagnetic waves underground through a transmitting antenna, and receives the electromagnetic waves reflected back to the ground through a receiving antenna. When the electromagnetic waves propagate in the underground medium, they are reflected when they encounter an interface with electrical differences. The spatial position, structure, morphology and burial depth of the underground medium are inferred based on the waveform, amplitude intensity and time change of the received electromagnetic waves. Ground penetrating radar has the advantages of fast detection speed, continuous detection process, high resolution, convenient and flexible operation, and low detection cost, and its application is gradually becoming more widespread. However, the images obtained by the current detection are mostly roughly identified by manual experience, which is inefficient and not very accurate.

[0005] In summary, how to achieve efficient and accurate recognition of images obtained by ground penetrating radar detection is an urgent problem to be solved. Summary of the invention

[0006] In view of this, it is necessary to provide an automatic recognition method for underground cable ground penetrating radar images based on similarity indicators to solve the urgent problem of how to achieve efficient and accurate recognition of images obtained by ground penetrating radar detection in the existing technology, thereby improving the efficiency and accuracy of ground penetrating radar in applications.

[0007] The present invention provides a method for automatically identifying underground cable ground penetrating radar images based on a similarity index, comprising the following steps:

[0008] Step S1: performing detection at a number of points randomly selected in a detection area by a ground penetrating radar to obtain a number of ground penetrating radar image data;

[0009] Step S2: Processing the ground penetrating radar image data to obtain a single-channel waveform under the electromagnetic wave radiation of the ground penetrating radar;

[0010] Step S3: Detect the area where the typical medium is located, collect the single-channel waveforms of each typical medium, and build a single-channel waveform library of the typical medium;

[0011] Step S4: calculating a similarity index between a single-channel waveform obtained by detecting an area of ​​unknown medium type by the ground penetrating radar and a waveform in a single-channel waveform library of the typical medium;

[0012] Step S5: Determine the medium type at the point in the detection area according to the size of the similarity index.

[0013] Furthermore, the step S2 specifically includes: using the average cancellation method to remove the direct wave of the clutter in the ground penetrating radar data; for removing the coupled wave of the antenna in the clutter, the antenna part is aimed at the space without objects to obtain the air signal at that time; then normal detection is performed, and the signal received during detection is subtracted from the air signal to obtain the coupled wave of the wire, so that the most accurate waveform information of the underground detection body can be displayed on the image section with the maximum resolution; due to the electromagnetic wave radiation of the ground penetrating radar, a certain intensity of electric field and magnetic field will be formed around the underground cable, and finally a single-channel waveform diagram of the corresponding position is obtained by calculation.

[0014] Furthermore, the typical media include metals, non-metals, common wires and cables.

[0015] Furthermore, in step S4, the DTW-Hausdorff distance is introduced as a similarity index to improve the recognition accuracy. The specific calculation process of the DTW-Hausdorff distance is as follows: First, the similarity index between the detected waveform and each curve in the waveform library is calculated. It is assumed that there are z types of curves in the database, and the DTW distance {D d1 ,D d2 ,…,D di ,…,D dz} and Hausdorff distance {D h1 ,D h2 ,…,D hj ,…,D hz}; Then, the means of the two data sets are normalized and the processed data are summed as the final DTW-Hausdorff distance.

[0016] Furthermore, in step S5, the sizes of the similarity indicators are compared, and the smaller the calculated DTW-Hausdorff distance is, the more similar the shapes of the two curves are. That is, the typical medium corresponding to the smallest DTW-Hausdorff distance is the medium type of the detection point.

[0017] Compared with the prior art, the method for automatically identifying underground cable ground penetrating radar images based on similarity index of the present invention has the following beneficial effects:

[0018] The present invention uses a ground penetrating radar to obtain image data of the detection area, processes the image data to obtain a single-channel waveform under the electromagnetic wave radiation of the ground penetrating radar, and finally calculates the similarity index with the waveform in the single-channel waveform library of the typical medium to determine the medium type of the detection area. Compared with traditional manual qualitative identification, a ground penetrating radar image recognition method for underground cables based on similarity index is provided, which reduces the cable identification errors that may be caused by human subjective factors, improves the objectivity and consistency of the results, and the calculation process is simple and convenient, and can process a large amount of data in a rapid and automated manner, avoiding the identification process that consumes human resources and time, and showing higher accuracy and stability on large-scale data. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic flow chart of a method for automatic recognition of underground cable ground penetrating radar images based on a similarity index provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings. The implementation scheme is described in detail below.

[0021] See also Figure 1 The embodiment of the present invention provides a method for automatically identifying underground cable ground penetrating radar images based on a similarity index, comprising the following steps:

[0022] Step S1: Detecting a number of points randomly selected in a detection area using a ground penetrating radar to obtain a number of ground penetrating radar image data.

[0023] Step S2: Since there are often a lot of clutter in the ground penetrating radar received signal, it is necessary to process the data collected by the ground penetrating radar to effectively suppress random and regular interference, as well as noise generated by internal and external factors.

[0024] In the clutter, direct waves and antenna coupling waves account for the largest proportion. According to the actual application of electricity, the average cancellation method is used as the primary method to remove direct waves. To remove the antenna coupling wave, the antenna part is aimed at the space without objects to obtain the air signal at that time; then normal detection is carried out, and the detection receiving signal at this time is subtracted from the air signal to obtain the coupling wave of the wire. Through the above processing, the image section can show the most accurate waveform information of the underground detection body with the maximum resolution. And because a certain intensity of electric field and magnetic field will be formed around the underground cable under the radiation of the ground penetrating radar electromagnetic wave, the single-channel waveform diagram of the corresponding position can be obtained by calculation.

[0025] Step S3: Detect the area where typical media such as metal, non-metallic media, ordinary wires and cables are located, collect the single-channel waveforms of each typical medium, and correspond them to each medium type. The typical medium single-channel waveform library finally obtained mainly includes the single-channel waveforms of typical metals, non-metallic media and ordinary wires and cables under the electromagnetic wave radiation of the ground penetrating radar.

[0026] Step S4: Calculate the similarity index between the single-channel waveform obtained by detecting the area of ​​unknown medium type by ground penetrating radar and the waveform in the single-channel waveform library of typical media. Dynamic Time Warping (DTW) distance and Hausdorff distance are two similarity evaluation indicators widely used in pattern recognition and artificial intelligence fields. Considering the properties of DTW distance, it can capture the local feature similarity between two groups of sequences. On the other hand, Hausdorff distance explains the overall differences within the data and can compare the similarity of overall shape and size. The present invention comprehensively utilizes the advantages of DTW distance and Hausdorff distance, and introduces DTW-Hausdorff distance as a similarity index to improve the accuracy of recognition.

[0027] The DTW algorithm is often used to compare the similarity of two time series by finding an optimal path between the two time series and aligning them on the time axis to quantify the degree of similarity between them.

[0028] Suppose there are two time series X and Y, with lengths n and m respectively:

[0029]

[0030] Use Euclidean distance as the distance metric between time series, and set x i and j The distance between them is denoted as d(x i ,y j ), where 1≤i≤n, 1≤j≤m. Then, create a cumulative distance matrix D, where D(i,j) represents the time step xi and j The cumulative distance between the two, the dimension of the matrix is ​​n×m. i With y j When aligned, it can be expressed as w k , so the final regular path W can be expressed as:

[0031] W={w1,w2,…,w k ,…,w K}max(n,m)≤K≤n+m-1 (2)

[0032] Before finding the optimal regular path, the corresponding boundary conditions need to be met:

[0033]

[0034] There are many paths from (1,1) to (n,m), but the optimal regular path we are looking for is the path with the minimum cumulative distance. Under the conditions of satisfying the monotonicity constraint and the continuity constraint, we can get the recursive formula for calculating the minimum cumulative distance as follows:

[0035]

[0036] Finally, the minimum cumulative distance D(n,m) between sequences X and Y can be obtained. D(n,m) can be used to judge the similarity of two sequences. The smaller its value is, the more similar the shapes of the two sequences are; and vice versa.

[0037] Hausdorff distance is a measure of the similarity between two sets of points. n} and Y={y1,y2,…,y m}, then the one-way Hausdorff distance between these two sets of data is:

[0038]

[0039] Where ∥·∥ is the Euclidean distance between point sets X and Y, and d h (X,Y),d h (Y,X) is the maximum value of the shortest distance from a point in X to the Y set and the maximum value of the shortest distance from a point in Y to the X set.

[0040] The bidirectional Hausdorff distance is shown in formula (6):

[0041] D h (X,Y)=max{d h (X,Y),d h (Y,X)} (6)

[0042] Where D h (X,Y) take D h (X,Y) and d h The maximum value in (Y, X). This distance measures the dissimilarity of two data sets and can be used to judge the similarity between X and Y. h The smaller the (X,Y) value, the more similar the X and Y shapes are.

[0043] The present invention makes use of the advantages of DTW distance and Hausdorff distance and introduces DTW-Hausdorff distance as a similarity index. This combination is intended to improve the accuracy of waveform recognition. The specific calculation process of DTW-Hausdorff distance is as follows:

[0044] First, the similarity index between the detected waveform and each curve in the waveform database is calculated. Assuming there are z types of curves in the database, this generates the DTW distance {D d1 ,D d2 ,…,D di ,…,D dz} and Hausdorff distance {D h1 ,D h2 ,…,D hj ,…,D hz Then, the means of the two data sets are standardized, and the calculation result of formula (9) is used as the final DTW-Hausdorff distance.

[0045]

[0046] D dhk =0.5·D dk ′+0.5·D hk ′,k=1,2,…,z (9)

[0047] Where D dhk As the DTW-Hausdorff distance between two data sets, the similarity between the detected waveform and the k-th waveform in the waveform library can be evaluated.

[0048] Step S5: Compare the sizes of the similarity indices. The smaller the calculated DTW-Hausdorff distance is, the more similar the shapes of the two curves are. That is, the typical medium corresponding to the smallest DTW-Hausdorff distance is the medium type of the detection point. Table 1 shows the similarity indices calculated from the waveforms obtained by randomly selecting three locations in the detection area, as well as the final determined medium type:

[0049] Table 1

[0050]

[0051] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for automatic recognition of underground cable ground penetrating radar images based on similarity index, characterized in that: include: Step S1: performing detection at a number of points randomly selected in a detection area by a ground penetrating radar to obtain a number of ground penetrating radar image data; Step S2: Processing the ground penetrating radar image data to obtain a single-channel waveform under the electromagnetic wave radiation of the ground penetrating radar; Step S3: Detect the area where the typical medium is located, collect the single-channel waveforms of each typical medium, and build a single-channel waveform library of the typical medium; Step S4: calculating a similarity index between a single-channel waveform obtained by detecting an area of ​​unknown medium type by the ground penetrating radar and a waveform in a single-channel waveform library of the typical medium; Step S5: Determine the medium type at the point in the detection area according to the size of the similarity index.

2. The method for automatic recognition of underground cable ground penetrating radar images based on similarity index according to claim 1 is characterized in that: The step S2 specifically includes: using the average cancellation method to remove the direct wave of the clutter in the ground penetrating radar data; for removing the antenna coupled wave in the clutter, the antenna part is aimed at the space without objects to obtain the air signal at that time; then normal detection is performed, and the signal received during detection is subtracted from the air signal to obtain the coupled wave of the wire, so that the most accurate waveform information of the underground detection body can be displayed on the image section with the maximum resolution; due to the electromagnetic wave radiation of the ground penetrating radar, a certain intensity of electric field and magnetic field will be formed around the underground cable, and finally a single-channel waveform diagram of the corresponding position is obtained by calculation.

3. The method for automatic recognition of underground cable ground penetrating radar images based on similarity index according to claim 1 is characterized in that: The typical media include metals, non-metals, common wires and cables.

4. The method for automatic recognition of underground cable ground penetrating radar images based on similarity index according to claim 1 is characterized in that: In step S4, the DTW-Hausdorff distance is introduced as a similarity index to improve the recognition accuracy. The specific calculation process of the DTW-Hausdorff distance is as follows: First, the similarity index between the detected waveform and each curve in the waveform library is calculated. It is assumed that there are z types of curves in the database, and the DTW distance {D d1 ,D d2 ,…,D di ,…,D dz } and Hausdorff distance {D h1 ,D h2 ,…,D hj ,…,D hz }; Then, the means of the two data sets are normalized and the processed data are summed as the final DTW-Hausdorff distance.

5. The method for automatic recognition of underground cable ground penetrating radar images based on similarity index according to claim 1 is characterized in that: In step S5, the sizes of the similarity indicators are compared. The smaller the calculated DTW-Hausdorff distance is, the more similar the shapes of the two curves are. That is, the typical medium corresponding to the smallest calculated DTW-Hausdorff distance is the medium type of the detection point.

Images