Three-dimensional ground penetrating radar map generation method

Abstract

The invention discloses a three-dimensional ground penetrating radar map generation method. The method comprises the following steps that S1, a three-dimensional ground penetrating radar is used for collecting selected road internal structure information to obtain original data; s2, preprocessing the obtained original data; s3, carrying out normalization processing and trilinear interpolation on the longitudinal section data of the preprocessed radar data; s4, generating a three-dimensional ground penetrating radar map, wherein the three-dimensional ground penetrating radar map at least comprises a single-channel oscillogram, a slice map, a longitudinal profile map and a three-dimensional map; and S5, based on the three-dimensional diagram, introducing a gradient enhanced adaptive perspective algorithm, realizing perspective of the three-dimensional diagram, and highlighting a disease area. According to the method, an effective and convenient standardized map is formulated in combination with data collected by the three-dimensional ground penetrating radar, the problems of section discretization and information discontinuity are solved, and through the generated three-dimensional perspective drawing, the visualization effect of a road disease area is effectively enhanced, and accurate positioning and recognition of road internal diseases are achieved.

Description

Technical Field

[0001] The present application relates to the field of ground penetrating radar detection technology, and in particular to a method for generating a three-dimensional ground penetrating radar map. Background Art

[0002] In recent years, highways have experienced a period of rapid construction, and many roads are approaching or exceeding their designed service life. Coupled with the impact of multiple factors such as long-term climate and vehicle load, road defects are gradually emerging, and highways have entered a comprehensive "maintenance era." Therefore, to extend road life and ensure traffic safety, a comprehensive and accurate assessment of road usage is necessary to determine maintenance plans.

[0003] Ground-penetrating radar (GPR), an efficient and nondestructive road inspection method, can visually detect internal defects such as voids, cracks in the base layer, and poor interlayer adhesion. It has been widely used to assess and diagnose pavement conditions. Traditional two-dimensional (2D) GPR uses a small number of antennas, acquiring only a single longitudinal profile, which fails to fully reflect the road's spatial information. Compared to traditional 2D radar, 3D GPR utilizes multiple antennas, enabling comprehensive road coverage. 3D GPR utilizes multiple detection channels, essentially generating multiple 2D longitudinal profiles in the depth direction. However, the resulting data still suffers from discrete spatial distribution and poor continuity, making it difficult to form a complete 3D model. This limits its effectiveness in fine-grained defect identification and 3D reconstruction. There is also a lack of a unified annotation process across image processing methods and algorithms. Furthermore, traditional methods struggle to visually display defect areas, and defects are not prominently displayed in 3D images. Summary of the Invention

[0004] In order to solve the above technical problems, an embodiment of the present application provides a three-dimensional ground penetrating radar map generation method, which enhances effective data information through a data preprocessing process, and adopts a trilinear interpolation method to solve the problems of discrete data distribution and poor continuity; based on the processed data, single-channel waveform diagrams, longitudinal profile diagrams, slice diagrams, three-dimensional diagrams, etc. are constructed to identify internal road surface diseases; and based on the three-dimensional diagram, a gradient enhancement adaptive perspective algorithm is introduced to realize the perspective of the three-dimensional diagram, highlight the diseased area, and realize the accurate positioning and identification of internal road diseases.

[0005] The present invention provides a method for generating a three-dimensional ground penetrating radar map, comprising the following steps:

[0006] Step S1: using a three-dimensional ground penetrating radar to collect the internal structure information of the selected road to obtain raw data;

[0007] Step S2: preprocessing the acquired raw data;

[0008] Step S3: normalizing and performing trilinear interpolation on the longitudinal profile data of the pre-processed radar data;

[0009] Step S4: Generate a three-dimensional ground penetrating radar map, which at least includes a single-channel waveform map, a slice map, a longitudinal profile map and a three-dimensional map.

[0010] Step S5: Based on the three-dimensional image, a gradient enhanced adaptive perspective algorithm is introduced to achieve perspective of the three-dimensional image and highlight the diseased area.

[0011] Furthermore, the step S2 includes at least one of the following:

[0012] Zero bias removal is used to remove the inherent DC low-frequency components generated by radar equipment and restore the antenna receiving signal;

[0013] Zero point adjustment is used to align the ground of each channel to eliminate the difference in the distance between the ground penetrating radar transmitting antenna and the ground when sampling starts between channels;

[0014] Gain adjustment is used to compensate for the attenuation of the detection medium, enhance the detection target signal, and strengthen the echo signal of deep-seated diseases;

[0015] Digital filtering, used to eliminate interference waves in the echo signal and extract useful signals from the echo signal;

[0016] Background elimination is used to remove invalid signals such as coupled pulses, surface reflections, and system ringing effects in radar data, eliminate horizontal oscillation stripes at fixed positions in the image, and highlight abnormal targets.

[0017] Furthermore, the zero bias removal specifically includes: , where Indicates the amplitude of the original single-channel echo signal without processing; It represents the amplitude of a single-channel echo signal after zero-bias processing; N represents the number of sampling points of a single-channel waveform, and n represents the nth sample of each single-channel echo signal scan.

[0018] Furthermore, the zero point adjustment specifically includes:

[0019] Find the time point t0 with the maximum amplitude;

[0020] Subtract this value from the data of each waveform to obtain the accurate starting position.

[0021] Furthermore, the gain adjustment specifically includes:

[0022] The energy attenuation method of the gain function is used for gain processing: , where a and b represent the linear gain coefficient and exponential gain coefficient respectively.

[0023] Furthermore, the digital filtering specifically includes:

[0024] Select 1 / 4 times the center frequency to 2 times the center frequency.

[0025] Furthermore, the background elimination specifically includes: , where Indicates the amplitude of the reflected echo signal after processing; It represents the amplitude of the original channel reflection echo signal before processing; k is the number of selected channels.

[0026] Furthermore, in step S3, the data of the 14 channels are normalized to [-1, 1], and the specific calculation formula is as follows: , where: Refers to the normalized value; Refers to the original data; and are the maximum and minimum values ​​in the data set, respectively.

[0027] Furthermore, in step S3, the normalized longitudinal profile data are interpolated at intervals of two to obtain a three-dimensional body of the road internal disease; the interpolation adopts trilinear interpolation.

[0028] Further, in step S4,

[0029] The single-channel waveform diagram is generated by plotting the depth as the ordinate and the reflection intensity as the abscissa, which reflects the relationship between the reflection intensity and depth of the underground medium interface;

[0030] The horizontal axis of the longitudinal cross-sectional view represents the position of the ground penetrating radar along the driving direction, the vertical axis represents the vertical depth downward, and the color or grayscale value represents the intensity of the reflected signal;

[0031] The slice diagram is a reflection signal distribution diagram at a specific depth within the road. Both the horizontal and vertical axes represent the road's planar coordinates, and the color or grayscale value represents the intensity of the reflection signal. Slice diagrams can intuitively reflect the spatial distribution of underground structures at that depth and are particularly suitable for analyzing the planar morphology, distribution range, and regional characteristics of defects, such as cracks.

[0032] The 3D map is a data fusion of longitudinal sections and slices (planar distribution at different depths) taken at multiple locations to reconstruct a three-dimensional model of the road's internal structure, known as a 3D radar map. This 3D radar map can fully and intuitively display the characteristics of road defects, including their spatial distribution, morphology, and specific location, providing a scientific visualization basis for road inspection and disease diagnosis.

[0033] Furthermore, in step S5, the process of realizing perspective of the three-dimensional image is as follows:

[0034] use The three-dimensional median filter extracts the background field, suppresses the reflection signal of the uniform medium, retains the abnormal disease characteristics, and constructs

[0035] Where: is the denoised 3D volume data; is the original three-dimensional volume data; It is a three-dimensional median filter operation;

[0036] Introducing the gradient enhanced adaptive perspective algorithm: , , where: is a gradient map, which represents the rate of change of three-dimensional data in the x, y, and z directions; 、 、 The gradients in the x, y, and z directions respectively indicate the rate of change of the data in these directions; is the enhanced three-dimensional data;

[0037] Construct an adaptive transfer function map to visualize a 3D plot:

[0038] , where: v represents ; ; ; For transparency.

[0039] Beneficial effects of the present invention: The present application provides a map generation algorithm that can combine data collected by three-dimensional ground penetrating radar and formulate effective and convenient standardized maps to solve the problems of discrete sections and discontinuous information. In addition, through the generated three-dimensional perspective map, it can effectively enhance the visualization effect of road disease areas and realize the precise positioning and identification of internal road diseases. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0041] Figure 1 It is a schematic diagram of the overall process of the present invention;

[0042] Figure 2 It is the data preprocessing process of the present invention;

[0043] Figure 3This is a comparison chart of the effects before and after zero bias removal in the present invention;

[0044] Figure 4 This is a comparison chart of the effects before and after zero point adjustment in the present invention;

[0045] Figure 5 This is a comparison diagram of the effects before and after gain adjustment in the present invention;

[0046] Figure 6 This is a comparison diagram of the effects before and after background elimination in the present invention;

[0047] Figure 7 Schematic diagram of the trilinear interpolation method of the present invention;

[0048] Figure 8 This is the step-by-step process of generating the three-dimensional ground penetrating radar map described in the present invention;

[0049] Figure 9 is a single-channel waveform diagram generated by an embodiment of the present invention;

[0050] Figure 10 is a longitudinal cross-sectional view generated by an embodiment of the present invention;

[0051] Figure 11 is a slice diagram generated by an embodiment of the present invention;

[0052] Figure 12 is a three-dimensional graph generated by an embodiment of the present invention;

[0053] Figure 13 It is a three-dimensional perspective image generated by an embodiment of the present invention. DETAILED DESCRIPTION

[0054] In order to make the purpose, features, and advantages of this application more obvious and easy to understand, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the embodiments described below are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0055] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0056] In the description of this application, it should be understood that the terms "upper", "lower", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on this application.

[0057] The present application is described below with reference to specific embodiments:

[0058] like Figure 1 、 2 As shown, the method for generating a three-dimensional ground penetrating radar spatial map involved in this application includes:

[0059] Step S1: Use three-dimensional ground penetrating radar to collect road internal structure information to obtain original data.

[0060] As a specific example, during on-site data collection, each lane is 3.75m wide, and the radar's effective measurement width is 1.82m. Therefore, two measurement lines are required to cover the entire lane. The collected raw data is output in .slc format, representing the two-way travel time of electromagnetic waves transmitted and received by each channel, along the road's travel direction. The 3D ground-penetrating radar has 14 channels (7 transmit and 14 receive). During data collection, the radar detects a maximum speed of 80km / h, a minimum sampling interval of 2.5cm, an antenna frequency of 1200MHz, a detection depth of 80cm, and a time window of 20ns. In this example, data from a section of the Qinglan Expressway is extracted, with 500 waveform data points per channel.

[0061] Step S2: Preprocess the detected radar raw data.

[0062] The processing method in this embodiment includes:

[0063] Zero bias removal, such as Figure 3 As shown, it is used to remove the inherent DC low-frequency component generated by the radar equipment and restore the antenna receiving signal.

[0064] As a specific embodiment, zero bias removal specifically includes: , where Indicates the amplitude of the original single-channel echo signal without processing; It represents the amplitude of a single-channel echo signal after zero-bias processing; N represents the number of sampling points of a single-channel waveform, and n represents the nth sample of each single-channel echo signal scan.

[0065] Zero point adjustment, such as Figure 4 As shown, the purpose is to align different channels on the data axis and to align the ground of each channel to eliminate the difference in the distance between the ground penetrating radar transmitting antenna and the ground when sampling starts between each channel.

[0066] As a specific example, the time point t0 at which the amplitude is maximum is found; this value is subtracted from the data of each waveform to obtain the accurate starting position. Traditional methods use multi-channel unified zero-point adjustment. However, due to differences in antenna hardware and uneven road surfaces, unified adjustment can introduce depth calculation errors, affecting the accuracy of 3D reconstruction. This application uses independent zero-point adjustment on a single channel to eliminate the inherent delay differences in the transmit / receive circuits of each channel and reduce the ground reflection time alignment error.

[0067] In this example, t0 is 2.73ns, which is given by the following formula: , where v is the propagation speed of electromagnetic waves in the medium, and its value is: , c represents the propagation speed of electromagnetic waves in vacuum. In this example, c=3×10 8 m / s; Indicates the relative dielectric constant of the medium. In this example, = 9. In this example, the position of the time window 0 is adjusted downward by 13.65 cm.

[0068] Gain adjustment, such as Figure 5 As shown, it is used to compensate for the attenuation of the detection medium, enhance the detection target signal, and strengthen the echo signal of deep-seated diseases; the energy attenuation method of the gain function is used for gain processing: , where a and b represent the linear gain coefficient and the exponential gain coefficient respectively. In this embodiment, the exponential gain form of 50% is selected.

[0069] Digital filtering is used to eliminate interference from the echo signal and extract the useful signal. Bandpass filtering is the most commonly used filtering method in ultra-wideband pulse ground-penetrating radar (UWB PGR). Theoretically, the bandwidth of a ground-penetrating radar is 1 / 2 to 3 / 2 times the center frequency. However, due to hardware and other limitations, the actual frequency band of the radar is slightly larger than this. Therefore, a range of 1 / 4 to 2 times the center frequency can also be selected.

[0070] In this embodiment, a linear phase finite impulse response (FIR) filter is used, 300 MHz is used for low-frequency signals, and 2400 MHz is used for high-frequency signals.

[0071] Background removal, such as Figure 6 As shown, it is used to remove invalid signals such as coupled pulses, surface reflections, and system ringing effects in radar data, and eliminate horizontal oscillation stripes at fixed positions in the image to highlight abnormal targets.

[0072] As a specific embodiment, the processing method is as follows: , where Indicates the amplitude of the reflected echo signal after processing; It represents the amplitude of the original channel reflection echo signal before processing; k is the number of selected channels.

[0073] Step S3: To reduce the deviation of the feature distribution of the data set and eliminate the influence of the difference in the subsequent interpolation dimension, the longitudinal profile data is normalized and trilinear interpolated.

[0074] In this example, the data of the 14 channels are normalized to [-1, 1]. The specific calculation formula is as follows: , where: Refers to the normalized value; Refers to the original data; and are the maximum and minimum values ​​in the data set, respectively.

[0075] In order to obtain continuous radar data, the normalized longitudinal profile data are interpolated at intervals of two to obtain the three-dimensional volume of the road internal disease. In this example, 4 sets of data are interpolated at intervals of two to obtain 66 sets of data. The specific interpolation method is trilinear interpolation, such as Figure 7 The specific calculation formula is as follows:

[0076]

[0077] Where:

[0078] :The point you are looking for The amplitude value,

[0079] :vertex The amplitude value of

[0080] :vertex The amplitude value of

[0081] :vertex The amplitude value of

[0082] :vertex The amplitude value of

[0083] :vertex The amplitude value of

[0084] :vertex The amplitude value of

[0085] :vertex The amplitude value of

[0086] :vertex Amplitude value.

[0087] Step S4: Generate a three-dimensional ground penetrating radar map based on the obtained data, such as Figure 8 As shown, it includes single-channel waveform diagram; slice diagram; longitudinal section diagram; three-dimensional diagram.

[0088] The single-channel waveform diagram is plotted with depth as the ordinate and reflection intensity as the abscissa; it reflects the relationship between the reflection intensity and depth of the underground medium interface.

[0089] The longitudinal profile is a projection of one-dimensional waveform data onto a two-dimensional plane. The horizontal axis represents the GPR's position along the travel direction, the vertical axis represents the vertical depth, and the color or grayscale value represents the intensity of the reflected signal. The longitudinal profile can reflect the longitudinal cross-sectional structure of the road from the surface to a certain depth.

[0090] The slice diagram is a reflection signal distribution diagram at a specific depth within the road. Both the horizontal and vertical axes represent the road's planar coordinates, and the color or grayscale value represents the intensity of the reflection signal. Slice diagrams can intuitively reflect the spatial distribution of underground structures at that depth and are particularly suitable for analyzing the planar morphology, distribution range, and regional characteristics of defects, such as cracks.

[0091] In this example, the single-channel waveform obtained by plotting the depth as the ordinate and the reflection intensity as the abscissa is as follows: Figure 9 shown.

[0092] In this example, the position of the ground penetrating radar along the driving direction is used as the horizontal axis, and the vertical depth is used as the vertical axis. The gray value in the figure represents the intensity of the reflected signal. The longitudinal cross-section is as follows: Figure 10 shown.

[0093] In this example, the position of the ground penetrating radar along the driving direction is used as the horizontal axis, and the vertical axis is perpendicular to the driving direction. The gray value in the figure represents the intensity of the reflected signal. The slice diagram is as follows: Figure 11 shown.

[0094] In this example, the longitudinal sections and slices (planar distribution at different depths) at multiple different locations are fused, and the position of the ground penetrating radar along the driving direction is used as the horizontal axis, the vertical axis perpendicular to the driving direction is used as the vertical axis, and the depth downward is used as the Z axis to reconstruct a three-dimensional image of the internal structure of the road, such as Figure 12 As shown in the figure, the gray value represents the intensity of the reflected signal.

[0095] The 3D map is a data fusion of longitudinal sections and slices (planar distribution at different depths) taken at multiple locations. This data can be used to reconstruct a three-dimensional model of the road's internal structure, known as a 3D radar map. This 3D radar map can fully and intuitively display the characteristics of road defects, including their spatial distribution, morphology, and specific location, providing a scientific visualization basis for road inspection and disease diagnosis.

[0096] Step S5: Based on the three-dimensional image, a gradient enhanced adaptive perspective algorithm is used to achieve perspective of the three-dimensional image and highlight the diseased area.

[0097] ① Background field suppression and noise elimination

[0098] A 5×5×5 three-dimensional median filter is used to extract the background field, suppress the uniform medium reflection signal, and retain the abnormal disease characteristics, as follows:

[0099]

[0100] Where: is the denoised 3D volume data; is the original three-dimensional volume data; It is a three-dimensional median filter operation.

[0101] ② Gradient domain disease edge enhancement

[0102]

[0103]

[0104] Where: It is a gradient map, which represents the rate of change of three-dimensional data in the x, y, and z directions; 、 、 The gradients in the x, y, and z directions respectively indicate the rate of change of the data in these directions; The enhanced three-dimensional data.

[0105] ③Adaptive transfer function mapping

[0106]

[0107] Where: v represents ; ; ; For transparency.

[0108] like Figure 13 What is shown is the perspective view of the three-dimensional image finally obtained in the embodiment of the present application, in which the internal defective area is enhanced and highlighted, so as to realize the accurate positioning and identification of internal defects of the road.

[0109] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0110] The preferred embodiments of the present invention are described in detail above, but the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, various equivalent transformations (such as quantity, shape, position, etc.) can be made to the technical solution of the present invention, and these equivalent transformations are all protected by the present invention.

Claims

1. A method for generating a three-dimensional ground penetrating radar map, characterized in that: The steps include: Step S1: using a three-dimensional ground penetrating radar to collect the internal structure information of the selected road to obtain raw data; Step S2: preprocessing the acquired raw data; Step S3: normalizing and performing trilinear interpolation on the longitudinal profile data of the pre-processed radar data; Step S4: generating a three-dimensional ground penetrating radar map, which at least includes a single-channel waveform map, a slice map, a longitudinal profile map, and a three-dimensional map; Step S5: Based on the three-dimensional image, a gradient enhanced adaptive perspective algorithm is introduced to achieve perspective of the three-dimensional image and highlight the diseased area.

2. The method for generating a three-dimensional ground penetrating radar map according to claim 1, wherein: The pre-processing in step S2 includes zero bias removal, zero point adjustment and gain adjustment performed in sequence; The zero bias removal is used to remove the inherent DC low-frequency component generated by the radar equipment, restore the antenna receiving signal, and avoid the introduction of baseline drift in subsequent steps; The zero point adjustment is used to align the ground of each channel to eliminate the difference in the distance between the ground penetrating radar transmitting antenna and the ground when sampling starts between each channel; The gain adjustment is used to compensate for the attenuation of the detection medium, enhance the detection target signal, and strengthen the echo signal of deep-seated diseases.

3. The method for generating a three-dimensional ground penetrating radar map according to claim 2, wherein: The pre-processing in step S2 further includes selective digital filtering and background elimination after gain adjustment; The digital filter is used to eliminate interference waves in the echo signal and extract useful signals from the echo signal; The background elimination is used to remove invalid signals such as coupled pulses, surface reflections, and system ringing effects in radar data, eliminate horizontal oscillation stripes at fixed positions in the image, and highlight abnormal targets.

4. The method for generating a three-dimensional ground penetrating radar map according to claim 3, wherein: The zero bias removal specifically includes: , where Represents the amplitude of the original single-channel echo signal before processing; Represents the amplitude of the processed single-channel echo signal; Indicates the number of sampling points of a single-channel waveform, and n represents the nth sample of each single-channel echo signal scan; The zero point adjustment specifically includes: Find the time point t0 with the maximum amplitude; subtract this value from the data of each waveform to obtain the accurate starting position; The gain adjustment specifically includes: The energy attenuation method of the gain function is used for gain processing: , where a and b represent the linear gain coefficient and exponential gain coefficient respectively.

5. The method for generating a three-dimensional ground penetrating radar map according to claim 4, wherein: The digital filtering specifically includes: The bandpass filtering method is used. Based on the bandwidth of the ground penetrating radar and the hardware limitations, the frequency band range is selected from 1 / 4 times the center frequency to 2 times the center frequency.

6. The method for generating a three-dimensional ground penetrating radar map according to claim 5, characterized in that: The background elimination specifically includes: , where Indicates the amplitude of the reflected echo signal after processing; It represents the amplitude of the original channel reflection echo signal before processing; k is the number of selected channels.

7. The method for generating a three-dimensional ground penetrating radar map according to claim 5, characterized in that: In step S3, the data of each channel is normalized to [-1, 1], and the formula is as follows: , where: Refers to the normalized value; Refers to the original data; and are the maximum and minimum values ​​in the data set, respectively.

8. The method for generating a three-dimensional ground penetrating radar map according to claim 7, wherein: In step S3, the normalized longitudinal profile data are interpolated at intervals of two to obtain a three-dimensional body of the road internal disease; the interpolation adopts trilinear interpolation.

9. The method for generating a three-dimensional ground penetrating radar map according to claim 8, wherein: In step S5, the process of realizing perspective of the three-dimensional image is as follows: use The three-dimensional median filter extracts the background field, suppresses the reflection signal of the uniform medium, retains the abnormal disease characteristics, and constructs , Where: is the denoised 3D volume data; is the original three-dimensional volume data; It is the three-dimensional median filter operation data; Introducing the gradient enhanced adaptive perspective algorithm: , , where: is a gradient map, which represents the rate of change of three-dimensional data in the x, y, and z directions; 、 、 The gradients in the x, y, and z directions respectively indicate the rate of change of the data in these directions; is the enhanced three-dimensional data; Construct an adaptive transfer function map to visualize a 3D plot: , where: v represents ; ; ; For transparency.

Images