Underground pipeline target identification method and device based on ground penetrating radar

Abstract

The invention relates to the technical field of underground pipeline detection, and discloses an underground pipeline target identification method and device based on a ground penetrating radar, and the method comprises the steps: collecting the multi-polarization scattering characteristics of an underground pipeline based on a multi-polarization radar or a single-stage radar, and obtaining the original image data of the underground pipeline; fusing the multi-polarization scattering characteristics by adopting a Laplacian pyramid algorithm to obtain fused image data; and solving a plurality of polarization attributes of the fused image data, and identifying a target pipeline in the underground pipelines based on a particle swarm AdaBoost algorithm and the polarization attributes. According to the method, the underground pipeline target is identified and classified by adopting a multi-polarization technology and combining with a deep learning method, so that the problem of low target classification accuracy caused by incomplete information of a single-polarization ground penetrating radar is solved, the target classification and identification speed is improved, and technical support is provided for further perfecting urban infrastructures.

Description

Technical Field

[0001] The present invention relates to the technical field of underground pipeline detection, and in particular to an underground pipeline target recognition method and device based on ground penetrating radar. Background Art

[0002] With the continuous acceleration of economic development and modernization, the scale of cities continues to expand. At the same time, the underground pipe network system as urban infrastructure has become more and more complex. However, due to the long time, some underground pipe network laying drawings cannot be traced, and in the process of urbanization, urban facilities are constantly updated, which brings about problems such as difficult maintenance of underground pipe network systems and frequent safety accidents. For example, when excavating and repairing underground pipe networks, accidents such as gas explosions and cable leakage may occur due to the loss or mismatch of the original drawings, which will bring great changes to residents' lives and even cause casualties. In order to reduce the incidence of such accidents, the detection and classification of underground pipeline systems is of great significance.

[0003] As a non-destructive detection method, ground penetrating radar has significant advantages such as high detection accuracy, high efficiency, high image resolution and easy operation. Ground penetrating radar obtains underground information by transmitting and receiving electromagnetic wave signals. When electromagnetic waves encounter targets with different electrical characteristics, their echo signals will show different characteristics. For underground pipeline systems, the materials, diameters and laying methods of water supply and drainage pipelines, gas pipelines, communication cables and power cables all have their own characteristics. Based on these characteristics, ground penetrating radar can be used to identify and classify underground pipeline targets, providing technical support for further improving urban infrastructure.

[0004] When facing targets with simple structures, traditional ground penetrating radars mostly use single polarization to identify targets by analyzing the amplitude, frequency, phase and other information of the echo signal. However, with the continuous improvement of urban infrastructure, the underground pipe network system is very complex, and single polarization ground penetrating radar cannot meet the detection needs. Multi-polarization technology can change the polarization characteristics of electromagnetic waves by changing the placement of the transceiver antenna, enrich the echo signal information, and analyze the polarization properties of the target to achieve the identification and classification of targets with complex distribution and structure. At the same time, in the classification process, it is necessary to extract features from the target information obtained by the ground penetrating radar. The feature extraction method that relies on manual extraction and shallow learning modules is not accurate and cannot meet the needs of underground pipeline systems with large data volumes. Summary of the invention

[0005] The present invention aims to at least solve the technical problem that in the classification process existing in the prior art, the feature extraction method relying on manual extraction and shallow learning modules is not accurate enough to meet the needs of underground pipeline systems with large data volumes.

[0006] To this end, an object of the present invention is to provide an underground pipeline target recognition method based on ground penetrating radar, comprising:

[0007] Based on the multi-polarization radar or the single-stage radar, the multi-polarization scattering characteristics of the underground pipeline are collected to obtain the original image data of the underground pipeline;

[0008] Using a Laplace pyramid algorithm to fuse the multi-polarization scattering characteristics to obtain fused image data;

[0009] Solve multiple polarization attributes of the fused image data, and identify the target pipeline in the underground pipeline based on the particle swarm AdaBoost algorithm and the polarization attributes, wherein the polarization attributes include polarization basic attributes, Feryman attributes, and polarization similarity attributes.

[0010] Furthermore, the multi-polarization scattering characteristics are fused using a Laplace pyramid algorithm to obtain fused image data, including:

[0011] Constructing a Gaussian pyramid with the original image data as the bottom layer;

[0012] Constructing a Laplace pyramid using the Gaussian pyramid, wherein the Laplace pyramid includes Laplace pyramids in multiple polarization directions;

[0013] The Laplacian pyramid is inverted to obtain the fused image data.

[0014] Furthermore, in the Gaussian pyramid, the image data of the Lth layer of the Gaussian pyramid is obtained by convolution calculation of the image data of the L-1th layer of the Gaussian pyramid and the low-pass filter window U, and the image data of the Lth layer of the Gaussian pyramid with a total number of layers Q is:

[0015]

[0016] Among them, i1 and j1 represent the number of rows and columns of data in the layer; m and n are the layer numbers; S L is the image data of the Lth layer of Gaussian pyramid in a single polarization direction; S L-1 is the image data of the L-1th layer of the Gaussian pyramid in a single polarization direction.

[0017] Further, constructing a Laplacian pyramid using the Gaussian pyramid includes:

[0018] Introduce operator A to interpolate and amplify the Gaussian pyramid to obtain S' L , wherein the S' L The size of is the same as the image data of the L-1th layer of the Gaussian pyramid, and the operator A is expressed as:

[0019]

[0020] Among them, i1 and j1 represent the number of rows and columns of data in the layer; m and n are the layer numbers; S L is the image data of the Lth layer of Gaussian pyramid in a single polarization direction; U represents the convolution of the low-pass filter window; S' L It is an intermediate value and has no real meaning;

[0021] Assume that the Lth layer Laplace pyramid data of a single polarization direction is P L , the formula is:

[0022] P L =S L -A(S L+1 )

[0023] Among them, S L+1 It is the image data of the L+1th layer of Gaussian pyramid in a single polarization direction; the Lth layer of Laplacian pyramid data is equal to the image data of the Lth layer of Gaussian pyramid.

[0024] Furthermore, the target pipeline in the underground pipeline is identified based on the particle swarm AdaBoost algorithm and the polarization attribute:

[0025] Construct a data set D based on the plurality of polarization attributes, and iteratively calculate the particle center PC of the data set D; establish a new data set Z based on the particle center PC, where Z = |D-PC|;

[0026] Project the data set D to the new domain and determine the weight of each sample point in the data set Z, where n k =(ω k1 ,ω k2 ,...,ω kl ,...,ω kN ), where k represents the iteration time; N represents the number of samples; ω kl is the weight of the lth sample point when the iteration time is k;

[0027] According to the weight and error curve E of each sample point k , calculate the error of the weak classifier as follows:

[0028]

[0029] Among them, e k is the error of the weak classifier; y l is the label of the lth sample point; Z l is the lth sample point in the new data set Z; h k is a weak classifier, and the weak classifier is based on the error curve E k Sure:

[0030]

[0031] Among them, r is the radius of the weak classifier;

[0032] According to the error of the weak classifier, the weight of the iterative sample point is updated, the underground pipeline is classified, and the target pipeline is identified.

[0033] Further, according to the error of the weak classifier, the weight of the iterative sample point is updated, including:

[0034] If the error of the weak classifier is less than the preset error threshold, it is a correct classification, otherwise it is an incorrect classification. In the next iteration, the weight of the correctly classified sample point is reduced, and the weight of the incorrectly classified sample point is increased.

[0035] Furthermore, the particle center PC of the data set D is calculated iteratively, including:

[0036] Calculate the speed and coordinates of each particle in the data set D at the iteration, where the speed of the qth particle in the i3th iteration is and coordinates They are:

[0037]

[0038] Among them, the subscript number of q represents the serial number of polarization attribute, 1 represents polarization entropy, 2 represents average scattering angle, 3-5 represents the three-dimensional parameter value of Freeman attribute, and 6-8 represents the three-dimensional parameter value used to characterize polarization similarity;

[0039] The loss function is constructed using the distance between the particle center and all particles, where the loss function is:

[0040]

[0041] Wherein, Loss is the loss function; N is the total number of particles; b is the particle number; c represents the number of polarization attributes, ranging from 1 to 8, 1 represents polarization entropy, 2 represents the average scattering angle, 3 to 5 represents the three-dimensional parameter value of Freeman attribute, and 6 to 8 represent the three-dimensional parameter values ​​used to characterize polarization similarity; D bc is the particle center when the sequence number of the polarization attribute in the data set D is c;

[0042] The minimum fitting value of the loss function of each particle from the 1st to the i3th iteration is calculated, and the minimum fitting value among all particles is determined as the particle center PC.

[0043] Furthermore, the method further comprises:

[0044] Update the velocity and coordinates of the particle group using the minimum fitting value among all particles;

[0045] The minimum fitting value is recalculated using the updated speed and coordinates of each particle until the difference between the minimum fitting value and the minimum fitting value in the previous iteration is less than a preset fitting value threshold.

[0046] Furthermore, the multi-polarization scattering characteristics of underground pipelines are collected based on multi-polarization radar or single-stage radar, including:

[0047] Using a multi-polarization radar and a chaotic signal as a detection signal to collect the multi-polarization scattering characteristics of the underground pipeline; or

[0048] A single-polarization radar is used, and the single-polarization radar is rotated at the same position to change the polarization direction of the antenna, collect scattering characteristics in multiple polarization directions, and obtain multi-polarization scattering characteristics of the underground pipeline.

[0049] The present invention provides an underground pipeline target recognition device based on ground penetrating radar, comprising: a collection module configured to collect multi-polarization scattering characteristics of underground pipelines based on multi-polarization radar or single-stage radar to obtain original image data of the underground pipelines;

[0050] A fusion module is configured to fuse the multi-polarization scattering characteristics using a Laplace pyramid algorithm to obtain fused image data;

[0051] The identification module is configured to solve multiple polarization attributes of the fused image data, and identify the target pipeline in the underground pipeline based on the particle swarm AdaBoost algorithm and the polarization attributes, wherein the polarization attributes include polarization basic attributes, Feryman attributes and polarization similarity attributes.

[0052] The basic properties of polarization include polarization entropy and mean scattering angle.

[0053] The present invention provides a method and device for identifying underground pipeline targets based on ground penetrating radar, which has the following beneficial effects:

[0054] (1) Obtaining multi-polarization information by changing the antenna position: Due to the continuous transformation in the process of urbanization, the underground pipe network system is very complex, and single-polarization information cannot meet the requirements for accurate identification of underground pipe network targets. Therefore, a multi-polarization ground penetrating radar is used to enrich target information and improve recognition and classification accuracy. However, due to the limitations of actual scenarios, there may be some special scenarios where it is inconvenient to use a multi-polarization radar and only a single-polarization radar can be used. To solve this problem, the present invention proposes that the direction of the single-polarization ground penetrating radar antenna can be continuously changed to obtain fully polarized target information under four relative position relationships.

[0055] (2) Using the Laplace pyramid algorithm to achieve multi-polarization data fusion: After obtaining multi-polarization target information based on ground penetrating radar, the image data needs to be multi-polarized and fused. When using the basic Gaussian pyramid, during the calculation process, the image will lose some high-frequency information after convolution and downsampling operations. The high-frequency information is used to describe the target detail information. When facing underground pipeline targets, due to the subtle differences in diameter and material between some pipes or cables, the Gaussian pyramid cannot meet the requirements for detail recognition. Therefore, the present invention uses the Laplace pyramid and the interpolation method to improve the sensitivity to detail information and avoid classification errors caused by information loss.

[0056] (3) Using the particle swarm AdaBoost method to achieve classification: Due to the complex position relationship of underground pipeline targets and low differentiation, multiple attribute values ​​need to be used as the classification basis during the classification process. Therefore, the particle swarm AdaBoost method is used to improve the classification accuracy. When facing multi-parameter problems, this method can automatically select appropriate parameters through weak classifiers according to different targets to improve classification efficiency.

[0057] The present invention adopts multi-polarization technology and combines it with deep learning methods to realize the recognition and classification of underground pipeline targets, solves the problem of incomplete information of single-polarization ground penetrating radar, resulting in low target classification accuracy, improves the target classification and recognition speed, and provides technical support for further improving urban infrastructure. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0059] Figure 1 It is a hybrid FDTD algorithm flow chart of an underground pipeline target recognition method based on ground penetrating radar in an embodiment of the present invention;

[0060] Figure 2 It is a schematic diagram of four position relationships between a transmitting antenna and a receiving antenna in a ground penetrating radar of an underground pipeline target recognition method based on a ground penetrating radar in an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0062] like Figure 1 As shown, a method for identifying underground pipeline targets based on ground penetrating radar according to an embodiment of the present disclosure includes the following steps:

[0063] Step S1: collecting multi-polarization scattering characteristics of the underground pipeline based on a multi-polarization radar or a single-stage radar to obtain original image data of the underground pipeline;

[0064] Step S2: using a Laplace pyramid algorithm to fuse the multi-polarization scattering characteristics to obtain fused image data;

[0065] Step S3: Solve multiple polarization attributes of the fused image data, and identify the target pipeline in the underground pipeline based on the particle swarm AdaBoost algorithm and the polarization attributes, wherein the polarization attributes include polarization basic attributes, Feryman attributes and polarization similarity attributes, and the polarization basic attributes include polarization entropy and average scattering angle.

[0066] Specifically, in step S1, when only a single-polarization radar can be used due to environmental reasons, multiple measurements are performed from multiple directions to obtain multi-polarization scattering characteristics.

[0067] More specifically, in step S1, a multi-polarization radar is used, and a chaotic signal is used as a detection signal to collect multi-polarization scattering characteristics. When a multi-polarization radar cannot be used in actual situations, a single-polarization radar is used, and the single-polarization radar is rotated at the same position to change the polarization direction of the antenna, thereby collecting and obtaining the multi-polarization scattering characteristics of the underground pipe network target, such as Figure 2 The figure shows the four positional relationships between the transmitting antenna and the receiving antenna in the ground penetrating radar, where TX represents the transmitting antenna, RX represents the receiving antenna, and the four relationships are horizontal transmission and horizontal reception (HH), vertical transmission and vertical reception (VV), horizontal transmission and vertical reception (HV), and vertical transmission and horizontal reception (VH).

[0068] The present invention proposes a method and device for identifying underground pipeline targets based on ground penetrating radar:

[0069] (1) Obtaining multi-polarization information by changing the antenna position: Due to the continuous transformation in the process of urbanization, the underground pipe network system is very complex, and single-polarization information cannot meet the requirements for accurate identification of underground pipe network targets. Therefore, a multi-polarization ground penetrating radar is used to enrich target information and improve recognition and classification accuracy. However, due to the limitations of actual scenarios, there may be some special scenarios where it is inconvenient to use a multi-polarization radar and only a single-polarization radar can be used. To solve this problem, the present invention proposes that the direction of the single-polarization ground penetrating radar antenna can be continuously changed to obtain fully polarized target information under four relative position relationships.

[0070] (2) Using the Laplace pyramid algorithm to achieve multi-polarization data fusion: After obtaining multi-polarization target information based on ground penetrating radar, the image data needs to be multi-polarized and fused. When using the basic Gaussian pyramid, during the calculation process, the image will lose some high-frequency information after convolution and downsampling operations. The high-frequency information is used to describe the target detail information. When facing underground pipeline targets, due to the subtle differences in diameter and material between some pipes or cables, the Gaussian pyramid cannot meet the requirements for detail recognition. Therefore, the present invention uses the Laplace pyramid and the interpolation method to improve the sensitivity to detail information and avoid classification errors caused by information loss.

[0071] (3) Using the particle swarm AdaBoost method to achieve classification: Due to the complex position relationship of underground pipeline targets and low differentiation, multiple attribute values ​​need to be used as the classification basis during the classification process. Therefore, the particle swarm AdaBoost method is used to improve the classification accuracy. When facing multi-parameter problems, this method can automatically select appropriate parameters through weak classifiers according to different targets to improve classification efficiency.

[0072] In one embodiment, after obtaining the multi-polarization scattering characteristics of the underground pipeline target, in order to improve the accuracy, step S2 in the underground pipeline target recognition method based on ground penetrating radar further includes the following steps:

[0073] This step includes three processes: decomposition, superposition and reconstruction. The specific process is as follows:

[0074] Step S21: construct a Gaussian pyramid with the original image data as the bottom layer, and use the convolution of the image data of the L-1th layer of the Gaussian pyramid and the low-pass filter window to obtain the image data of the Lth layer of the Gaussian pyramid with a total number of layers Q:

[0075]

[0076] Among them, i1 and j1 represent the number of rows and columns of data in this layer; m and n are the layer numbers, and U(m,n) is the low-pass filter window of this layer; S L is the image data of the Lth layer of the Gaussian pyramid in a certain polarization direction; SL-1 is the image data of the L-1th layer of the Gaussian pyramid in a certain polarization direction;

[0077] Step S22: obtaining Laplacian pyramid data using the image data of the Gaussian pyramid, wherein the Laplacian pyramid includes Laplacian pyramids in multiple polarization directions;

[0078] Specifically, the operator A is introduced to interpolate and amplify the Gaussian pyramid to obtain S' L , wherein the S' L The size of is the same as the image data of the L-1th layer of the Gaussian pyramid, and the operator A is expressed as:

[0079]

[0080] Among them, i1 and j1 represent the number of rows and columns of data in the layer; m and n are the layer numbers; S L is the image data of the Lth layer of Gaussian pyramid in a single polarization direction; U represents the convolution of the low-pass filter window; S' L It is an intermediate value and has no real meaning;

[0081] Assume that the Lth layer Laplace pyramid data of a single polarization direction is P L , the formula is:

[0082] P L =S L -A(S L+1 )

[0083] Among them, S L+1 It is the image data of the L+1th layer of Gaussian pyramid in a single polarization direction; the Lth layer of Laplacian pyramid data is equal to the image data of the Lth layer of Gaussian pyramid.

[0084] Step S23: Invert the Laplacian pyramid to obtain the fused image data, including the horizontally transmitted and horizontally received Laplacian pyramid image data. Laplacian pyramid data for horizontal transmission and vertical reception and Laplace pyramid data of vertical transmission and horizontal reception

[0085] S L-1 =P L-1 -A(S L )

[0086] In the image data of the Laplacian pyramid, a specific threshold is taken to divide the high-frequency area and the low-frequency area. The high frequency describes the detail information, and the low frequency describes the contour information. Therefore, for the high-frequency part, the maximum value is taken; for the low-frequency part, the average value is taken.

[0087] In step S3, the polarization basic properties, the Freeman properties, and the polarization similar properties include polarization entropy H, average scattering angle α, and three-dimensional parameter values ​​P for characterizing the Freeman properties. S , P d , P v and the three-dimensional parameter value rs used to characterize the polarization similarity s ,rs v rs d ;

[0088] The above attributes are obtained from the fused image data based on theoretical formulas, where the subscripts s, v, and d represent scattering, double reflection scattering, and volume scattering, respectively;

[0089] In one embodiment, in step S3 of the underground pipeline target identification method based on ground penetrating radar, identifying the target pipeline in the underground pipeline based on the particle swarm AdaBoost algorithm and the polarization attribute further includes the following steps:

[0090] Step S31: constructing a data set D based on the plurality of polarization attributes, and calculating the particle center PC of the data set D by iteration; establishing a new data set Z according to the particle center PC, wherein Z=|D-PC|;

[0091] Step S32: Project the data set D to a new domain and determine the weight of each sample point in the data set Z, where n k =(ω k1 ,ω k2 ,...,ω kl ,...,ω kN ), where k represents the iteration time; N represents the number of samples; ω kl is the weight of the lth sample point when the iteration time is k;

[0092] Step S33: Based on the weight of each sample point and the error curve E k , calculate the error of the weak classifier;

[0093] Specifically, as follows:

[0094]

[0095] Among them, e k is the error of the weak classifier; y l is the label of the lth sample point; Z l is the lth sample point in the new data set Z; h k is a weak classifier, and the weak classifier is based on the error curve E k Sure:

[0096]

[0097] Where r is the radius of the weak classifier.

[0098] Step S34: According to the error of the weak classifier, the weight of the iterative sample point is updated, the underground pipeline is classified, and the target pipeline is identified. If the error of the weak classifier is less than the preset error threshold, it is a correct classification, otherwise it is an incorrect classification. In the next iteration, the weight of the correctly classified sample point is reduced, and the weight of the incorrectly classified sample point is increased.

[0099] In step S31, calculating the particle center PC of the data set D includes the following steps:

[0100] Step S311: Calculate the speed and coordinates of each particle in the data set D during iteration, where the speed of the qth particle in the i3th iteration is and coordinates They are:

[0101]

[0102] Among them, the subscript number of q represents the serial number of the polarization attribute, 1 represents polarization entropy, 2 represents the average scattering angle, 3-5 represents the three-dimensional parameter value of the Freeman attribute, and 6-8 represents the three-dimensional parameter value used to characterize the polarization similarity; the details are as follows:

[0103] is the velocity value of the qth particle of the polarization entropy H in the i3th iteration; is the velocity value of the qth particle with average scattering angle α in the i3th iteration; is the scattering parameter value P used to characterize the Freeman property S The velocity value of the qth particle in the i3th iteration; is the double-reflection scattering parameter value P used to characterize the Freeman property d The velocity value of the qth particle in the i3th iteration; The volume scattering parameter value P used to characterize the Freeman property v The velocity value of the qth particle in the i3th iteration; is the scattering parameter value rs used to characterize the polarization similarity s The velocity value of the qth particle in the i3th iteration; The double reflection scattering parameter value rs used to characterize the polarization similarity v The velocity value of the qth particle in the i3th iteration; The volume scattering parameter value rs used to characterize the polarization similarity d The velocity value of the qth particle in the i3th iteration;

[0104] is the coordinate value of the qth particle of the polarization entropy H in the i3th iteration; is the coordinate value of the qth particle with average scattering angle α in the i3th iteration; is the scattering parameter value P used to characterize the Freeman property S The coordinate value of the qth particle in the i3th iteration; is the double-reflection scattering parameter value P used to characterize the Freeman property d The coordinate value of the qth particle in the i3th iteration; The volume scattering parameter value P used to characterize the Freeman property v The coordinate value of the qth particle in the i3th iteration; is the scattering parameter value rs used to characterize the polarization similarity s The coordinate value of the qth particle in the i3th iteration; The double reflection scattering parameter value rs used to characterize the polarization similarity v The coordinate value of the qth particle in the i3th iteration; The volume scattering parameter value rs used to characterize the polarization similarity d The coordinate value of the qth particle in the i3th iteration;

[0105] Step S312: Using the distance between the particle center PC and all particles, a loss function is constructed as follows:

[0106]

[0107] Wherein, Loss is the loss function; N is the total number of particles; b is the particle number; c represents the number of polarization attributes, ranging from 1 to 8, 1 represents polarization entropy, 2 represents the average scattering angle, 3 to 5 represents the three-dimensional parameter value of Freeman attribute, and 6 to 8 represent the three-dimensional parameter values ​​used to characterize polarization similarity; D bc is the particle center when the sequence number of the polarization attribute in the data set D is c;

[0108] Step S313: Calculate the minimum fitting value of the loss function of each particle from the 1st to the i3th iteration, and determine the minimum fitting value among all particles as the particle center PC;

[0109] Update the velocity and coordinates of the particle group using the minimum fitting value among all particles;

[0110] The minimum fitting value is recalculated using the updated speed and coordinates of each particle until the difference between the minimum fitting value and the minimum fitting value in the previous iteration is less than a preset fitting value threshold.

[0111] The details are as follows:

[0112] Calculate the minimum fitting value of the particle from the 1st to the i3th iteration. The minimum fitting value of the qth particle is The minimum fitting value among all particles is PC i3 , the calculation formula is as follows:

[0113]

[0114] The disclosed embodiment also provides an underground pipeline target recognition device based on a ground penetrating radar, comprising: a collection module configured to collect multi-polarization scattering characteristics of the underground pipeline based on a multi-polarization radar or a single-stage radar to obtain original image data of the underground pipeline;

[0115] A fusion module is configured to fuse the multi-polarization scattering characteristics using a Laplace pyramid algorithm to obtain fused image data;

[0116] The identification module is configured to solve multiple polarization attributes of the fused image data, and identify the target pipeline in the underground pipeline based on the particle swarm AdaBoost algorithm and the polarization attributes, wherein the polarization attributes include polarization basic attributes, Feryman attributes and polarization similarity attributes.

[0117] The embodiment of the present disclosure further provides an electronic device, comprising at least a memory and a processor, wherein the memory stores a computer program, and the processor implements the above-mentioned underground pipeline target identification method based on ground penetrating radar when executing the computer program on the memory, and the method comprises the following steps:

[0118] Step S1: collecting multi-polarization scattering characteristics of the underground pipeline based on a multi-polarization radar or a single-stage radar to obtain original image data of the underground pipeline;

[0119] Step S2: using a Laplace pyramid algorithm to fuse the multi-polarization scattering characteristics to obtain fused image data;

[0120] Step S3: Solve multiple polarization attributes of the fused image data, and identify the target pipeline in the underground pipeline based on the particle swarm AdaBoost algorithm and the polarization attributes, wherein the polarization attributes include polarization basic attributes, Feryman attributes and polarization similarity attributes, and the polarization basic attributes include polarization entropy and average scattering angle.

[0121] In some embodiments, the processor executing the computer program may be a processing device including one or more general-purpose processing devices, such as a microprocessor, a central processing unit (CPU), a graphics processing unit (GPU), etc. More specifically, the processor may be a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, a processor running other instruction sets, or a processor running a combination of instruction sets. The processor may also be one or more special-purpose processing devices, such as an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), a system on a chip (SoC), etc.

[0122] The memory may be a read-only memory (ROM), a random access memory (RAM), a phase-change random access memory (PRAM), a static random access memory (SRAM), a dynamic random access memory (DRAM), an electrically erasable programmable read-only memory (EEPROM), other types of random access memory (RAM), a flash disk or other form of flash memory, a cache, a register, a static memory, a compact disk read-only memory (CD-ROM), a digital versatile disk (DVD) or other optical storage, a cassette or other magnetic storage device, or any other possible non-temporary medium used to store information or instructions that can be accessed by a computer device.

[0123] The electronic devices of the embodiments of the present disclosure may include but are not limited to fixed terminal devices such as MCU controllers, servers, desktop computers, digital TVs, etc., and mobile terminal devices such as vehicle-mounted devices (such as head-up display devices), handheld devices (such as mobile phones, tablet computers, etc.), wearable devices (such as smart watches, smart bracelets, etc.), etc.

[0124] The present disclosure also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the above-mentioned underground pipeline target recognition method based on ground penetrating radar is implemented. The method includes the following steps:

[0125] Step S1: collecting multi-polarization scattering characteristics of the underground pipeline based on a multi-polarization radar or a single-stage radar to obtain original image data of the underground pipeline;

[0126] Step S2: using a Laplace pyramid algorithm to fuse the multi-polarization scattering characteristics to obtain fused image data;

[0127] Step S3: Solve multiple polarization attributes of the fused image data, and identify the target pipeline in the underground pipeline based on the particle swarm AdaBoost algorithm and the polarization attributes, wherein the polarization attributes include polarization basic attributes, Feryman attributes and polarization similarity attributes, and the polarization basic attributes include polarization entropy and average scattering angle.

[0128] The present invention has the following beneficial effects:

[0129] (1) Obtaining multi-polarization information by changing the antenna position: Due to the continuous transformation in the process of urbanization, the underground pipe network system is very complex, and single-polarization information cannot meet the requirements for accurate identification of underground pipe network targets. Therefore, a multi-polarization ground penetrating radar is used to enrich target information and improve recognition and classification accuracy. However, due to the limitations of actual scenarios, there may be some special scenarios where it is inconvenient to use a multi-polarization radar and only a single-polarization radar can be used. To solve this problem, the present invention proposes that the direction of the single-polarization ground penetrating radar antenna can be continuously changed to obtain fully polarized target information under four relative position relationships.

[0130] (2) Using the Laplace pyramid algorithm to achieve multi-polarization data fusion: After obtaining multi-polarization target information based on ground penetrating radar, the image data needs to be multi-polarized and fused. When using the basic Gaussian pyramid, during the calculation process, the image will lose some high-frequency information after convolution and downsampling operations. The high-frequency information is used to describe the target detail information. When facing underground pipeline targets, due to the subtle differences in diameter and material between some pipes or cables, the Gaussian pyramid cannot meet the requirements for detail recognition. Therefore, the present invention uses the Laplace pyramid and the interpolation method to improve the sensitivity to detail information and avoid classification errors caused by information loss.

[0131] (3) Using the particle swarm AdaBoost method to achieve classification: Due to the complex position relationship of underground pipeline targets and low differentiation, multiple attribute values ​​need to be used as the classification basis during the classification process. Therefore, the particle swarm AdaBoost method is used to improve the classification accuracy. When facing multi-parameter problems, this method can automatically select appropriate parameters through weak classifiers according to different targets to improve classification efficiency.

[0132] The present invention adopts multi-polarization technology and combines it with deep learning methods to realize the recognition and classification of underground pipeline targets, solves the problem of incomplete information of single-polarization ground penetrating radar, resulting in low target classification accuracy, improves the target classification and recognition speed, and provides technical support for further improving urban infrastructure.

[0133] The computer-readable storage medium of the embodiments of the present disclosure may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. In the embodiments of the present disclosure, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device, for example, the memory described above.

[0134] The computer programs of the embodiments of the present disclosure may be organized into one or more computer executable components or modules. Any number and combination of such components or modules may be used to implement the various aspects of the present disclosure. For example, the various aspects of the present disclosure are not limited to the specific computer executable instructions or specific components or modules shown in the drawings and described herein. Other embodiments may include different computer executable instructions or components with more or less functions than those shown and described herein.

[0135] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

Claims

1. A method for identifying underground pipeline targets based on ground penetrating radar, characterized in that: include: Based on the multi-polarization radar or the single-stage radar, the multi-polarization scattering characteristics of the underground pipeline are collected to obtain the original image data of the underground pipeline; Using a Laplace pyramid algorithm to fuse the multi-polarization scattering characteristics to obtain fused image data; Solve multiple polarization attributes of the fused image data, and identify the target pipeline in the underground pipeline based on the particle swarm AdaBoost algorithm and the polarization attributes, wherein the polarization attributes include polarization basic attributes, Feryman attributes, and polarization similarity attributes.

2. The underground pipeline target identification method based on ground penetrating radar according to claim 1 is characterized in that: The multi-polarization scattering characteristics are fused using a Laplace pyramid algorithm to obtain fused image data, including: Constructing a Gaussian pyramid with the original image data as the bottom layer; Constructing a Laplace pyramid using the Gaussian pyramid, wherein the Laplace pyramid includes Laplace pyramids in multiple polarization directions; The Laplacian pyramid is inverted to obtain the fused image data.

3. The underground pipeline target identification method based on ground penetrating radar according to claim 2 is characterized in that: In the Gaussian pyramid, the image data of the Lth layer of the Gaussian pyramid is obtained by convolution calculation of the image data of the L-1th layer of the Gaussian pyramid and the low-pass filter window U. The image data of the Lth layer of the Gaussian pyramid with a total number of layers Q is: Among them, i1 and j1 represent the number of rows and columns of data in the layer; m and n are the layer numbers; S L is the image data of the Lth layer of Gaussian pyramid in a single polarization direction; S L-1 is the image data of the L-1th layer of the Gaussian pyramid in a single polarization direction.

4. The underground pipeline target identification method based on ground penetrating radar according to claim 2 is characterized in that: Using the Gaussian pyramid to construct a Laplacian pyramid includes: Introduce operator A to interpolate and amplify the Gaussian pyramid to obtain S' L , wherein the S' L The size of is the same as the image data of the L-1th layer of the Gaussian pyramid, and the operator A is expressed as: Among them, i1 and j1 represent the number of rows and columns of data in the layer; m and n are the layer numbers; S L is the image data of the Lth layer of Gaussian pyramid in a single polarization direction; U represents the convolution of the low-pass filter window; S' L It is an intermediate value and has no real meaning; Assume that the Lth layer Laplace pyramid data of a single polarization direction is P L , the formula is: P L =S L -A(S L+1 ) Among them, S L+1 It is the image data of the L+1th layer of Gaussian pyramid in a single polarization direction; the Lth layer of Laplacian pyramid data is equal to the image data of the Lth layer of Gaussian pyramid.

5. The underground pipeline target identification method based on ground penetrating radar according to claim 1 is characterized in that: Identify the target pipeline in the underground pipeline based on the particle swarm AdaBoost algorithm and the polarization attribute: Construct a data set D based on the plurality of polarization attributes, and iteratively calculate the particle center PC of the data set D; establish a new data set Z based on the particle center PC, where Z = |D-PC|; Project the data set D to the new domain and determine the weight of each sample point in the data set Z, where n k =(ω k1 ,ω k2 ,...,ω kl ,...,ω kN ), where k represents the iteration time; N represents the number of samples; ω kl is the weight of the lth sample point when the iteration time is k; According to the weight and error curve E of each sample point k , calculate the error of the weak classifier as follows: Among them, e k is the error of the weak classifier; y l is the label of the lth sample point; Z l is the lth sample point in the new data set Z; h k is a weak classifier, and the weak classifier is based on the error curve E k Sure: Among them, r is the radius of the weak classifier; According to the error of the weak classifier, the weight of the iterative sample point is updated, the underground pipeline is classified, and the target pipeline is identified.

6. The underground pipeline target identification method based on ground penetrating radar according to claim 5 is characterized in that: According to the error of the weak classifier, the weight of the iterative sample point is updated, including: If the error of the weak classifier is less than the preset error threshold, it is a correct classification, otherwise it is an incorrect classification. In the next iteration, the weight of the correctly classified sample point is reduced, and the weight of the incorrectly classified sample point is increased.

7. The underground pipeline target identification method based on ground penetrating radar according to claim 5 is characterized in that: The particle center PC of the data set D is calculated iteratively, including: Calculate the speed and coordinates of each particle in the data set D at the iteration, where the speed of the qth particle in the i3th iteration is and coordinates They are: Among them, the subscript number of q represents the serial number of polarization attribute, 1 represents polarization entropy, 2 represents average scattering angle, 3-5 represents the three-dimensional parameter value of Freeman attribute, and 6-8 represents the three-dimensional parameter value used to characterize polarization similarity; The loss function is constructed using the distance between the particle center and all particles, where the loss function is: Wherein, Loss is the loss function; N is the total number of particles; b is the particle number; c represents the number of polarization attributes, ranging from 1 to 8, 1 represents polarization entropy, 2 represents the average scattering angle, 3 to 5 represents the three-dimensional parameter value of Freeman attribute, and 6 to 8 represent the three-dimensional parameter values ​​used to characterize polarization similarity; D bc is the particle center when the sequence number of the polarization attribute in the data set D is c; The minimum fitting value of the loss function of each particle from the 1st to the i3th iteration is calculated, and the minimum fitting value among all particles is determined as the particle center PC.

8. The underground pipeline target recognition method based on ground penetrating radar according to claim 7, the method further comprising: Update the velocity and coordinates of the particle group using the minimum fitting value among all particles; The minimum fitting value is recalculated using the updated speed and coordinates of each particle until the difference between the minimum fitting value and the minimum fitting value in the previous iteration is less than a preset fitting value threshold.

9. The underground pipeline target identification method based on ground penetrating radar according to claim 1 is characterized in that: The multi-polarization scattering characteristics of underground pipelines are collected based on multi-polarization radar or single-stage radar, including: Using a multi-polarization radar and a chaotic signal as a detection signal to collect the multi-polarization scattering characteristics of the underground pipeline; or A single-polarization radar is used, and the single-polarization radar is rotated at the same position to change the polarization direction of the antenna, collect scattering characteristics in multiple polarization directions, and obtain multi-polarization scattering characteristics of the underground pipeline.

10. An underground pipeline target identification device based on ground penetrating radar, characterized in that: include: A collection module, configured to collect multi-polarization scattering characteristics of underground pipelines based on multi-polarization radar or single-stage radar to obtain original image data of the underground pipelines; A fusion module is configured to fuse the multi-polarization scattering characteristics using a Laplace pyramid algorithm to obtain fused image data; The identification module is configured to solve multiple polarization attributes of the fused image data, and identify the target pipeline in the underground pipeline based on the particle swarm AdaBoost algorithm and the polarization attributes, wherein the polarization attributes include polarization basic attributes, Feryman attributes and polarization similarity attributes.

Images