A method and system for splicing and processing multi-channel data of three-dimensional ground penetrating radar

Abstract

The present invention relates to a method and system for splicing and processing multi-channel data of a three-dimensional ground penetrating radar, and belongs to the field of pavement structure disease detection. The method comprises: slicing the three-dimensional radar data to obtain XY plane slice data; performing latitude and longitude calculation on the data points in the XY plane slice data to obtain the latitude and longitude coordinates of all data points; determining the reference channel and using the adjacent channel judgment algorithm to find all adjacent channels of the reference channel; using the canvas range determination algorithm to determine the canvas range; determining the data point range of the reference channel and the adjacent channel in the canvas respectively according to the canvas range; interpolating, coordinate translating and rotating the data points within the data point range, mapping the data points to pixel points, and generating an image after the channel horizontal plane is spliced. Based on the image after the channel horizontal plane is spliced ​​generated by the method of the present invention, crack disease recognition can reduce the difficulty of crack disease recognition and improve the accuracy of crack disease recognition.

Description

Technical Field

[0001] The present invention relates to the technical field of pavement structure disease detection, and in particular to a method and system for splicing and processing multi-channel data of a three-dimensional ground penetrating radar. Background Art

[0002] Under the influence of driving load, temperature, ultraviolet rays, etc., the road pavement structure will inevitably suffer damage as the operation time increases, such as cracks, subsidence, interlayer voids, etc. Quickly, effectively and accurately find the location and type of pavement structure diseases and determine their severity, and predict the development of diseases, which is of great significance for scientifically and rationally determining the maintenance time and formulating maintenance plans. Among the various types of pavement structure diseases, crack diseases are the most numerous and have a greater impact on the service performance of pavement structures.

[0003] Ground penetrating radar detection technology is one of the main representative technologies for non-destructive testing of roads. It mainly obtains relevant information about the detected object by analyzing the propagation of electromagnetic waves inside the detected object. In terms of internal disease detection of old roads, it is similar to the "CT examination of the human body" to conduct visual diagnosis of the health status of the road, and realize rapid inspection of the quality of the road structure without a lot of core drilling work; compared with traditional detection technology, it has technical advantages such as non-destructive and high detection efficiency.

[0004] Among various ground penetrating radars, the two-dimensional ground penetrating radar has a low precision and accuracy in detecting cracks due to its single antenna setting. The three-dimensional ground penetrating radar collects multi-channel radar data at the same time to form three-dimensional radar data. The multi-channel joint judgment of crack damage reduces the misjudgment rate to a certain extent and is an effective means to detect crack damage inside the pavement structure.

[0005] However, due to the large spatial span of crack diseases, the features usually span multiple lanes. In the interpretation of 3D GPR crack disease images, single-channel 3D GPR data is currently widely used to identify crack diseases, which has the problems of narrow auxiliary judgment area, high identification difficulty, and certain misjudgment. Therefore, it is necessary to develop a 3D GPR data splicing method and pre-process the 3D GPR data to further improve the accuracy of crack disease identification. Summary of the invention

[0006] The purpose of the present invention is to provide a method and system for splicing and processing three-dimensional ground penetrating radar multi-channel data, so as to reduce the difficulty of crack disease identification and improve the accuracy of crack disease identification.

[0007] To achieve the above object, the present invention provides the following solutions:

[0008] A method for splicing and processing multi-channel data of a three-dimensional ground penetrating radar, comprising:

[0009] Acquire three-dimensional radar data of multiple measurement channels collected by a three-dimensional ground penetrating radar multi-channel antenna array and perform slicing processing on the three-dimensional radar data to obtain XY plane slicing data of the three-dimensional radar data;

[0010] Calculate the longitude and latitude of the data points in the XY plane slice data to obtain the longitude and latitude coordinates of all the data points;

[0011] Determine a reference track among the multiple tracks and use an adjacent track determination algorithm to find all adjacent tracks of the reference track;

[0012] Determine the canvas range using a canvas range determination algorithm according to the latitude and longitude coordinates of the data points in the reference survey track;

[0013] Determine the data point ranges of the reference track and the adjacent track in the canvas respectively according to the canvas range;

[0014] After interpolation, coordinate translation and rotation are performed on the data points within the data point range, the data points are mapped to pixel points to generate a spliced ​​image of the horizontal plane of the survey track.

[0015] Optionally, the calculating the longitude and latitude of the data points in the XY plane slice data to obtain the longitude and latitude coordinates of all the data points specifically includes:

[0016] The ground plane is approximated as a Euclidean plane, and the longitude and latitude coordinates of all data points are derived according to the longitude and latitude coordinates of the center of the multi-channel antenna array, the angle between the driving direction and the true north direction, and the distances between other data points and the central measuring point in the XY plane slice data. All data points are projected onto the satellite map according to the longitude and latitude coordinates of all data points.

[0017] Optionally, the step of using an adjacent track determination algorithm to find all adjacent tracks of the reference track specifically includes:

[0018] Obtaining the coordinate maximum points of the data points in the reference track and the track to be determined;

[0019] The first area is marked with the coordinate maximum point in the reference track as the diagonal point, and the second area is marked with the coordinate maximum point in the to-be-determined track as the diagonal point;

[0020] Calculate the ratio of the intersection of the first area and the second area to the union of the first area and the second area;

[0021] If the ratio is greater than a preset ratio threshold, it is determined that the to-be-determined measurement track is an adjacent measurement track of the reference measurement track.

[0022] Optionally, the determining the canvas range by using a canvas range determination algorithm according to the latitude and longitude coordinates of the data points in the reference survey track specifically includes:

[0023] Convert the latitude and longitude coordinates of the data points in the reference track into UTM coordinates, and set the center axis of the canvas to be consistent with the center axis of the data points in the middle track;

[0024] Selecting the starting boundary point and reference data point of the canvas from the data points of the first channel in the reference track;

[0025] Determine the UTM coordinates of the end boundary point of the canvas according to the UTM coordinates of the start boundary point and the reference data point;

[0026] The boundary line equation of the canvas range is determined according to the UTM coordinates of the starting boundary point, the reference data point and the ending boundary point.

[0027] Optionally, the determining the data point ranges of the reference track and the adjacent track in the canvas respectively according to the canvas range specifically includes:

[0028] Determining the data point range of the reference track in the canvas according to the boundary line equation of the canvas range; the boundary line of the canvas range includes a starting boundary line and an ending boundary line;

[0029] Sequentially detect the distance between each data point in the adjacent measurement track and the canvas boundary line, determine the data point with the shortest distance from the starting boundary line as the starting point, and determine the data point with the shortest distance from the ending boundary line as the ending point;

[0030] The data point range of the adjacent measurement track in the canvas is determined according to the coordinate range of the starting point and the ending point.

[0031] Optionally, after interpolating, translating and rotating the data points within the data point range, mapping the data points to pixel points to generate a spliced ​​image of the horizontal plane of the measurement track, specifically includes:

[0032] Mapping the data points within the data point range onto the canvas;

[0033] Use bilinear interpolation to interpolate the data points on the canvas;

[0034] After interpolation, the UTM coordinates of the data points on the canvas are translated, and the midpoint of the canvas is translated to the origin of the coordinates;

[0035] After translating the coordinates, rotate the coordinate system with the center of the canvas as the origin;

[0036] The data points after the rotation of the coordinate system are mapped to the image pixels, and the images of the horizontal planes of the measurement tracks are output one by one.

[0037] A three-dimensional ground penetrating radar multi-channel data splicing and processing system, comprising:

[0038] A data acquisition and preprocessing module, used to acquire three-dimensional radar data of multiple measurement channels collected by a three-dimensional ground penetrating radar multi-channel antenna array and slice the three-dimensional radar data to obtain XY plane slice data of the three-dimensional radar data;

[0039] A data point longitude and latitude calculation module is used to calculate the longitude and latitude of the data points in the XY plane slice data to obtain the longitude and latitude coordinates of all data points;

[0040] An adjacent channel determination module, used to determine a reference channel among the multiple channels and to find all adjacent channels of the reference channel using an adjacent channel determination algorithm;

[0041] A drawing canvas range determination module, used to determine the canvas range using a canvas range determination algorithm according to the latitude and longitude coordinates of the data points in the reference survey track;

[0042] A data point range determination module, used to determine the data point ranges of the reference track and the adjacent track in the canvas respectively according to the canvas range;

[0043] The track data point stitching module is used to interpolate, translate and rotate the data points within the data point range, map the data points to pixel points, and generate a stitched image of the track horizontal plane.

[0044] Optionally, the data point longitude and latitude calculation module specifically includes:

[0045] The data point longitude and latitude calculation unit is used to approximate the ground plane to the Euclidean plane, derive the longitude and latitude coordinates of all data points according to the longitude and latitude coordinates of the center of the multi-channel antenna array, the angle between the driving direction and the north direction, and the distance between other data points and the central measuring point in the XY plane slice data, and project all data points onto the satellite map according to the longitude and latitude coordinates of all data points.

[0046] Optionally, the adjacent measurement track judgment module specifically includes:

[0047] A coordinate maximum value point acquisition unit, used to acquire the coordinate maximum value points of the data points in the reference measurement track and the measurement track to be determined;

[0048] An area marking unit, used for marking a first area with the coordinate maximum point in the reference track as a diagonal point, and marking a second area with the coordinate maximum point in the to-be-determined track as a diagonal point;

[0049] a ratio calculation unit, used to calculate the ratio of the intersection of the first area and the second area to the union of the first area and the second area;

[0050] The adjacent channel determination unit is used to determine that the channel to be determined is an adjacent channel of the reference channel if the ratio is greater than a preset ratio threshold.

[0051] Optionally, the drawing canvas range determination module specifically includes:

[0052] A coordinate conversion unit, used to convert the latitude and longitude coordinates of the data points in the reference survey track into UTM coordinates, and set the central axis of the canvas to be consistent with the central axis of the data points in the intermediate survey track;

[0053] A starting boundary point selection unit, used to select a starting boundary point and a reference data point of the canvas from the data points of the first channel in the reference measurement track;

[0054] An end boundary point determination unit, used to determine the UTM coordinates of the end boundary point of the canvas according to the UTM coordinates of the start boundary point and the reference data point;

[0055] A canvas range boundary line determination unit is used to determine the boundary line equation of the canvas range according to the UTM coordinates of the starting boundary point, the reference data point and the ending boundary point.

[0056] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0057] The present invention provides a method and system for splicing and processing multi-channel data of a three-dimensional ground penetrating radar, the method comprising: obtaining three-dimensional radar data of multiple channels collected by a three-dimensional ground penetrating radar multi-channel antenna array and slicing the three-dimensional radar data to obtain XY plane slice data of the three-dimensional radar data; performing latitude and longitude calculation on the data points in the XY plane slice data to obtain the latitude and longitude coordinates of all data points; determining a reference channel among the multiple channels and using an adjacent channel judgment algorithm to find all adjacent channels of the reference channel; using a canvas range determination algorithm to determine the canvas range according to the latitude and longitude coordinates of the data points in the reference channel; respectively determining the data point range of the reference channel and the adjacent channel in the canvas according to the canvas range; interpolating, coordinate translating and rotating the data points within the data point range, mapping the data points to pixel points, and generating an image after the channel horizontal plane is spliced. Crack disease recognition based on the image after the channel horizontal plane is spliced ​​generated by the method of the present invention can reduce the difficulty of crack disease recognition and improve the accuracy of crack disease recognition.

[0058] In addition, when the image after splicing the horizontal plane of the survey channels generated by the method of the present invention is manually interpreted or software recognized and interpreted, multi-channel three-dimensional ground penetrating radar data located in the same cross section can be processed at one time, thereby greatly improving the efficiency of crack disease interpretation of radar data. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] 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 will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0060] Figure 1 It is a flow chart of a method for splicing and processing multi-channel data of three-dimensional ground penetrating radar according to the present invention;

[0061] Figure 2 A schematic diagram of a three-dimensional ground penetrating radar track arrangement provided in an embodiment of the present invention;

[0062] Figure 3 A schematic diagram of a vehicle-mounted ground penetrating radar detection site provided by an embodiment of the present invention;

[0063] Figure 4 Schematic diagram of three slicing methods of three-dimensional ground penetrating radar data provided by an embodiment of the present invention;

[0064] Figure 5 A schematic diagram of three slice images of three-dimensional ground penetrating radar data provided by an embodiment of the present invention;

[0065] Figure 6 A schematic diagram of the sampling positions of data points and the actual recorded longitude and latitude positions provided by an embodiment of the present invention;

[0066] Figure 7 A schematic diagram of the longitude and latitude coordinates of derived data points provided by an embodiment of the present invention;

[0067] Figure 8 A schematic diagram of a single-track XY section of a three-dimensional ground penetrating radar projected onto a satellite map provided by an embodiment of the present invention;

[0068] Fig. 9 A schematic diagram of determining whether two measurement tracks are adjacent to each other provided by an embodiment of the present invention;

[0069] Fig.10 A schematic diagram of determining a canvas range provided by an embodiment of the present invention;

[0070] Fig.11 A schematic diagram of determining a range of data points to be mapped onto a canvas provided by an embodiment of the present invention;

[0071] Fig.12 A schematic diagram of bilinear interpolation provided by an embodiment of the present invention;

[0072] Fig.13 A schematic diagram of translating a midpoint of a canvas to a coordinate origin provided by an embodiment of the present invention;

[0073] Fig.14 A schematic diagram of a coordinate system rotation process provided by an embodiment of the present invention;

[0074] Fig.15 A schematic diagram of an image after stitching five horizontal planes provided in an embodiment of the present invention;

[0075] Fig.16 A schematic diagram of a single-channel three-dimensional ground penetrating radar image provided by an embodiment of the present invention;

[0076] Fig.17 The present invention is a flowchart of a specific embodiment of a method for splicing and processing three-dimensional ground penetrating radar multi-channel data. DETAILED DESCRIPTION

[0077] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0078] The purpose of the present invention is to provide a method and system for splicing and processing three-dimensional ground penetrating radar multi-channel data, so as to reduce the difficulty of crack disease identification and improve the accuracy of crack disease identification.

[0079] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0080] Figure 1 This is a flow chart of a method for splicing and processing 3D ground penetrating radar multi-channel data in the present invention. Figure 1 The present invention provides a method for splicing and processing three-dimensional ground penetrating radar multi-channel data, comprising:

[0081] Step 101: Acquire three-dimensional radar data of multiple channels collected by a three-dimensional ground penetrating radar multi-channel antenna array and perform slicing processing on the three-dimensional radar data to obtain XY plane slice data of the three-dimensional radar data.

[0082] First, data collection and preprocessing are performed. The multi-channel antenna array of the vehicle-mounted 3D ground penetrating radar has a scanning width of 1.5 meters each time. By setting up multiple detection paths, full-section coverage scanning of roads of any width can be achieved. The 3D ground penetrating radar measurement path is arranged as follows: Figure 2 On-site operation is to guide the driving trajectory of the inspection vehicle by making marks at equal distances (1.5m) at the starting point or using signs and markings on the road, such as Figure 3 As shown in Figure 1, the multi-channel antenna array of the vehicle-mounted 3D ground penetrating radar is used to collect 3D radar data of multiple measurement tracks.

[0083] In the process of radar data recognition and processing, the three-dimensional radar data is converted into two-dimensional image data in the form of slices for recognition and processing. First, an XYZ three-dimensional coordinate system is established, with the direction of the radar antenna as the X-axis, the depth direction perpendicular to the earth as the Z-axis, and the direction perpendicular to the XZ plane as the Y-axis. Then, the three-dimensional radar data is sliced ​​to obtain three mutually perpendicular slices, such as Figure 4 As shown, they are XY plane slices, YZ plane slices, and XZ plane slices respectively.

[0084] The single-channel crack signal is expressed in three mutually perpendicular slice signals as follows: Figure 5 As shown. It can be seen that the crack signal is not obvious in the YZ plane and the XZ plane, but is more obvious in the XY plane. This is related to the fact that the crack disease is long in length, but the width and height are generally small. Therefore, the method of the present invention mainly intercepts the XY plane slice data of the three-dimensional radar data to identify the crack disease.

[0085] Step 102: Calculate the longitude and latitude of the data points in the XY plane slice data to obtain the longitude and latitude coordinates of all data points.

[0086] The ground plane is approximated as a Euclidean plane. The longitude and latitude coordinates of all data points can be derived based on the longitude and latitude coordinates of the center of the multi-channel antenna array, the angle between the driving direction and the north direction, and the distances between other data points and the central measuring point in the XY plane slice data. All data points can then be projected onto the satellite map based on their longitude and latitude coordinates.

[0087] Specifically, since the 3D GPR data only records the longitude and latitude data of the center line of the survey data, to obtain the longitude and latitude data of all survey points (data points), it is necessary to calculate the longitude and latitude coordinates of each channel data point based on the longitude and latitude coordinates of the center of the survey track and the driving direction. Because during data collection, the location information recorded in the navigation file is the longitude and latitude coordinate value of the center of the antenna array, and the longitude and latitude coordinates of each sampled data point are not recorded. The line connecting the sampling positions of each data point is perpendicular to the driving direction, such as Figure 6As shown in the figure, the hollow circle is the actual position of the recorded longitude and latitude values, the solid circle is the actual sampling position of each data point, and the angle θ is the angle between the driving direction and the true north direction.

[0088] Since the distance between the location of each data point and the longitude and latitude coordinates is within 1m, at this accuracy, we can assume that the ground plane is approximately the Euclidean plane, and then derive the longitude and latitude coordinates of the data point. Figure 7 As shown, point A is the recorded longitude and latitude coordinates, point B is the longitude and latitude coordinates of the data point being sought, and point B is located southeast of point A. The longitude and latitude coordinates between AB satisfy the following formula:

[0089]

[0090]

[0091] Among them, lon1 and lat1 are the longitude and latitude coordinates of point A, lon2 and lat2 are the longitude and latitude coordinates of point B; γ is the complementary angle of θ; ||AB|| is the distance between point A and point B; R0=6371.393km is the average radius of the earth.

[0092] After calculating the longitude and latitude coordinates of all data points according to formulas (1) and (2), they can be projected onto the satellite map, as shown in Figure 8 shown.

[0093] Step 103: Determine a reference track among the multiple tracks and use an adjacent track determination algorithm to find all adjacent tracks of the reference track.

[0094] A certain section of the multiple channels is designated as the reference channel. After the reference channel is given, it is necessary to find the channel data adjacent to the current reference channel. Therefore, the present invention proposes an adjacent channel judgment algorithm. The method for judging whether two channels are adjacent is as follows: Fig. 9 As shown. First, find the coordinate extreme points of the data points in the two sections of the measurement track (reference measurement track and the measurement track to be determined), and mark the first area S1 and the second area S2 with the coordinate extreme points. Then, use formula (3) to calculate the ratio r of the intersection of areas S1 and S2 to the union of S1 and S2.

[0095] If the r value is greater than the specified threshold, the two sections are considered to be adjacent.

[0096]

[0097] Therefore, the step 103 uses an adjacent track judgment algorithm to find all adjacent tracks of the reference track, which specifically includes:

[0098] Step 3.1: Obtain the coordinate extreme points of the data points in the reference track and the track to be determined, the coordinate extreme points include the maximum coordinate point and the minimum coordinate point; Fig. 9 As shown, (x min1， y min1 ) and (x max1 ,y max1 ) are the coordinate values ​​of the maximum and minimum points in the reference track respectively; (x min2 ,y min2 ) and (x max2 ,y max2 ) are the coordinate values ​​of the maximum point and the minimum point in the path to be determined;

[0099] Step 3.2: Mark the first area S1 with the coordinate maximum point in the reference track as the diagonal point, and mark the second area S2 with the coordinate maximum point in the to-be-determined track as the diagonal point; wherein the coordinates of the diagonal points of the first area S1 are (x min1 ,y min1 ) and (x max1 ,y max1 ), the coordinates of the diagonal points of the second area S2 are (x min2 ,y min2 ) and (x max2 ,y max2 );

[0100] Step 3.3: Calculate the ratio r of the intersection S1∩S2 of the first region S1 and the second region S2 to the union S1∪S2 of the first region S1 and the second region S2 using formula (3);

[0101] Step 3.4: If the ratio r is greater than a preset ratio threshold, determine that the to-be-determined measurement track is an adjacent measurement track of the reference measurement track.

[0102] Step 104: Determine the canvas range using a canvas range determination algorithm according to the latitude and longitude coordinates of the data points in the reference survey track.

[0103] In order to improve the consistency of the stitched images, the center axis of the canvas for drawing the horizontal stitched images is set to be consistent with the center axis of the middle track data. After selecting multiple adjacent track data along the road width, the track in the middle is used as the reference track. The width is set according to the width of the detection area, and the length and width are kept consistent to determine the range of the canvas in the actual space. The process of determining the range of the canvas is as follows: Fig.10 shown.

[0104] The step 104 determines the canvas range using a canvas range determination algorithm according to the latitude and longitude coordinates of the data points in the reference survey track, and specifically includes:

[0105] Step 4.1: Convert the longitude and latitude coordinates of the data points in the reference track into UTM (Universal Transverse Mercator Grid System) coordinates, and set the central axis of the canvas to be consistent with the central axis of the data points in the middle track;

[0106] Step 4.2: Selecting the starting boundary point and reference data point of the canvas from the data points of the first channel in the reference measurement track;

[0107] After converting the latitude and longitude coordinates of the data point into UTM coordinates, select the data point of the first channel in the benchmark survey track and use point p i1 As the starting boundary point of the canvas. Then, select another point p in the data points of channel 1. j1 As a reference data point, the subscript j satisfies:

[0108]

[0109] Among them, the above is the set length of the canvas, ||p i1 p j1 || is the starting boundary point p i1 With reference data point p j1 The distance between them, arg min indicates the minimum value.

[0110] Step 4.3: Determine the UTM coordinates of the end boundary point of the canvas according to the UTM coordinates of the start boundary point and the reference data point;

[0111] Assume that p i1 The UTM coordinates are (x1, y1), p j1 The UTM coordinates are (x2, y2), then the end boundary point p of the canvas edge0 The UTM coordinates (x edge0 ,y edge0 ) can be obtained using formulas (5) and (6):

[0112]

[0113]

[0114] Step 4.4: Determine the boundary line equation of the canvas range according to the UTM coordinates of the starting boundary point, the reference data point, and the ending boundary point.

[0115] like Fig.10 As shown, assume that the straight line l0 is formed by p i1 and p j1 Determine the straight line, straight line l1 is perpendicular to l0 and passes through point p i1The straight line l2 is perpendicular to l0 and passes through point p edge0 The equations of l1 and l2 are shown in formulas (7) and (8):

[0116] l1:(x1-x2)x+(y1-y2)y+(y2-y1)y1+(x2-x1)x1=0 (7)

[0117] l2:(x1-x2)x+(y1-y2)y+(y2-y1)y edge0 +(x2-x1)x edge0 =0 (8)

[0118] Where l1 and l2 are the starting boundary line and the ending boundary line of the canvas range respectively.

[0119] The canvas range can be determined according to the canvas range boundary lines l1 and l2 and the set canvas width W, such as Fig.10 As shown, p edge1 、p edge2 、p edge3 、p edge4 are the four corner points of the canvas, p medium is the midpoint of the canvas, α is the angle between l0 and the x-axis of the UTM coordinate, and its tangent is the slope of l0.

[0120] Step 105: Determine the data point ranges of the reference track and the adjacent track in the canvas respectively according to the canvas range.

[0121] The step 105 determines the data point ranges of the reference track and the adjacent track in the canvas respectively according to the canvas range, and specifically includes:

[0122] Step 5.1: Determine the data point range of the reference track in the canvas according to the boundary line equation of the canvas range; Fig.11 The canvas range shown is the data point range of the reference track in the canvas. edge1 、p edge2 、p edge3 、p edge4 They are the four corner points of the data point range;

[0123] Step 5.2: sequentially detect the distance between each data point in the adjacent measurement channel and the canvas boundary line, and determine the data point with the smallest distance to the starting boundary line l1 as the starting point p begin , determine the data point with the smallest distance l2 from the termination boundary line as the termination point p end ;

[0124] Step 5.3: According to the starting point p begin With the termination point p endThe coordinate range determines the data point range of the adjacent track in the canvas.

[0125] The actual spatial range of the canvas determined by the reference track currently only contains the data points of the reference track, so it is necessary to find the data points in other tracks that fall into the canvas range. For the reference track, the subscript range corresponding to the data points that can be mapped to the canvas range is [i, k]. For other adjacent tracks except the reference track, to determine the data point subscript range, it is necessary to select the first channel, and then calculate the distance from each data point of the channel to l1 and l2, such as Fig.11 As shown. Among them, p begin is the starting point of the data, and the distance from this point to l1 is the smallest; p end is the end point of the data, and the distance from this point to l2 is the smallest. Therefore, the subscript range of the data points of the current adjacent channel is [begin, end], that is, the data point range of the adjacent channel in the canvas is obtained.

[0126] Step 106: After interpolation, coordinate translation and rotation of the data points within the data point range, the data points are mapped to pixel points to generate a spliced ​​image of the horizontal plane of the survey track.

[0127] The step 106 interpolates, translates and rotates the data points within the data point range, maps the data points to pixel points, and generates a spliced ​​image of the horizontal plane of the measurement track, which specifically includes:

[0128] Step 6.1: Mapping the data points within the data point range onto the canvas;

[0129] Step 6.2: Use bilinear interpolation to interpolate the data points on the canvas;

[0130] The original radar data is a discrete data dot matrix. After the data points are mapped to the canvas, the data points on the canvas will be unevenly distributed. There are two main reasons: first, when the data points are directly mapped to the canvas with uniform spatial distribution, the density of data points along the length and width of the road is different; second, in the subsequent coordinate system rotation process, some data points will overlap, resulting in uneven distribution of data points. In order to improve the quality of the generated image, the present invention uses bilinear interpolation to interpolate the data points to improve the continuity of the data points in space. The interpolation calculation method is as follows: Fig.12 For any adjacent detection data points P1, P2, P3, P4, insert a point Q at any position in the rectangular area with these four points as corner points. The function value calculation formula of this point is:

[0131]

[0132] Among them, (x1, y1, f1), (x1, y2, f2), (x2, y2, f3), (x2, y1, f4), and (x, y, f) are the xyz-axis coordinates of points P1, P2, P3, P4, and Q respectively.

[0133] Based on formula (9), the data value at any coordinate can be calculated and the data can be encrypted and interpolated.

[0134] Step 6.3: After interpolation, translate the UTM coordinates of the data points on the canvas and translate the midpoint of the canvas to the origin of the coordinates;

[0135] The UTM coordinates of the data points that fall within the canvas range are translated so that the coordinate system can be rotated with the center of the canvas as the origin. Fig.13 As shown, the midpoint p of the canvas needs to be medium (x medium ,y medium ) is moved to the origin of the coordinate system. Therefore, the relationship between the translated coordinates (x′, y′) and the original coordinates (x, y) is shown in formula (10):

[0136]

[0137] Among them, the midpoint p of the canvas medium The UTM coordinates are:

[0138]

[0139] Step 6.4: After translating the coordinates, rotate the coordinate system with the center of the canvas as the origin;

[0140] Since the direction of each road section is different, in order to unify the display effect of the generated image, after translating the coordinates, it is necessary to rotate the UTM coordinate system to correct the direction of the canvas. The rotation process is as follows Fig.14 As shown, where angle β is the angle between the original coordinate system and the rotated coordinate system. After the coordinate system is rotated counterclockwise by angle β, the relationship between the coordinate value (x″, y″) and the coordinate value (x′, y′) in the original coordinate system is shown in formula (12):

[0141]

[0142] Step 6.5: Map the data points after the coordinate system is rotated to the image pixels, and output the stitched images of the horizontal plane of the measurement track one by one.

[0143] The process of mapping the data points in the canvas to the image pixels is as follows: first, the canvas is discretized, and each coordinate point in the canvas corresponds to a pixel point after discretization; then, according to the order of the measurement tracks, the data point values ​​in the canvas are assigned to the pixels with the closest coordinates, and the spliced ​​images are output one by one, that is, the image after the horizontal plane of the output measurement tracks is spliced, such as Fig.15 shown.

[0144] Fig.16 The data shown is a single-channel 3D ground penetrating radar data. Fig.15 It can be seen from the three-dimensional ground penetrating radar rendering after the five channels are spliced ​​that it is difficult to identify and determine the crack area with a single channel data only relying on local information, while the image after multi-channel splicing processing using the method of the present invention can identify the crack area relying on global information, the difficulty of identifying crack diseases is greatly reduced, and the determination accuracy is greatly improved.

[0145] Fig.17 FIG. 1 is a flowchart of a specific embodiment of a method for splicing and processing multi-channel data of a three-dimensional ground penetrating radar according to the present invention. Fig.17 As shown, first, a certain section of the track is designated as the reference track, and the adjacent track judgment algorithm is used to find all adjacent tracks of the current reference track. Then, the longitude and latitude coordinate values ​​of all data points of the reference track and adjacent tracks are calculated through the input navigation file (3D radar data). Processing starts from the first data point of the first channel of the reference track, and the canvas range determination algorithm is used to determine the range of the current canvas. Next, the data point ranges of the reference track and adjacent tracks in the canvas are determined respectively. After data point interpolation, coordinate translation and rotation, the data points are mapped to pixel points, and finally an image of the track horizontal plane splicing is generated. The algorithm runs in a loop until all data points of the current reference track are processed.

[0146] Specifically, during data collection, the navigation file records the location information of the longitude and latitude coordinates of the center of the antenna array, and does not record the longitude and latitude coordinates of each sampling data point. It is assumed that the ground plane is approximately the Euclidean plane, and the longitude and latitude coordinates of all radar data points are derived based on the longitude and latitude coordinates of the center of the antenna array, the angle between the driving direction and the north direction, and the distance between other data points and the center point.

[0147] In order to ensure that the radar image drawn after stitching is not too large to consume computing resources or too small to cause radar data loss, the radar image canvas needs to be set to an appropriate position, direction and size. First, the longitude and latitude coordinates of all data points are converted into UTM coordinates, and the central axis of the canvas is set to be consistent with the central axis of the middle track data. The direction and position of the canvas are determined, and the width is set according to the width of the detection area. The length and width are kept consistent to determine the range of the canvas in actual space.

[0148] After a reference track is given, the ratio of the intersection of the reference track and the area marked by the coordinate maximum points of the adjacent track is calculated; if the ratio is greater than a specified threshold, the two sections of road are judged to be adjacent. The radar data adjacent track judgment algorithm proposed by the present invention determines the adjacent tracks belonging to the same section, which is the basis for subsequent splicing processing.

[0149] The actual spatial range of the canvas determined by the reference channel currently only contains the data points of the reference channel, so it is necessary to find the data points in other channels that fall within the canvas range. The distances of each point in the adjacent channels from the edge of the canvas are detected in turn, and the point with the smallest distance and less than a certain threshold is the starting point / end point; after determining the range of the measurement points, the measurement points within the range are interpolated, translated, and rotated to form standardized multi-channel splicing data, which is then projected onto the canvas to form a standard radar image of multi-channel splicing.

[0150] The method of the present invention is applied to the manual interpretation and software identification and interpretation of crack defects in radar data. The same road section is selected for comparison with the traditional single-track identification process. The statistical results are shown in Table 1 below.

[0151] Table 1

[0152]

[0153] The software identification and interpretation used in this experiment is the 3D ground penetrating radar crack disease identification software based on Crack Unet. The test shows that the interpretation efficiency is improved by more than 10 times when the method of the present invention is applied to the manual interpretation process, and the interpretation efficiency is improved by 2.5 times when the method is applied to the software identification and interpretation process. The accuracy is greatly improved in both experiments.

[0154] Since the single-channel 3D ground penetrating radar data is generally 1-2 meters wide, 8-30 channels are set, and the width of the road is generally 4-10 meters, multiple channels need to be set to cover the entire section of the road to be detected. At present, when judging the internal crack disease of the road, the 3D ground penetrating radar data is often processed one by one to identify the crack disease in the data, without fully considering other data in the same section. When the crack length is long and spans multiple channels, a single-channel 3D ground penetrating radar is used to judge the crack disease, which has the problems of narrow auxiliary judgment area, great difficulty in identification, and inaccurate crack area. The present invention proposes a 3D ground penetrating radar multi-channel data splicing processing method, which can realize the automatic splicing of multi-channel 3D ground penetrating radar data and output the spliced ​​images one by one, so as to realize the multi-channel comprehensive judgment of whether there is crack disease inside the pavement structure, and to judge the direction, length and other indicators of the crack disease, reduce the difficulty of crack disease identification, and improve the accuracy of the judgment of crack disease inside the structure.

[0155] Radar data processing is difficult. If manual interpretation is used, the interpretation process is time-consuming, the workload is huge, and the interpretation efficiency is low. If software recognition and interpretation are used, the software running time is long. Since 100% accuracy cannot be guaranteed, the interpretation results still need to be manually reviewed after the recognition is completed. According to the previous method of interpreting data one by one, it takes a lot of energy and time. If the method of splicing interpretation is adopted, the data of the same section can be processed at the same time, the workload can be reduced by more than 85%, and the interpretation efficiency can be greatly improved. The spliced ​​image generated by the method of the present invention can be applied to both manual interpretation of radar data and software recognition interpretation. It can process multi-channel three-dimensional ground penetrating radar data located in the same cross section at one time, so it can greatly improve the efficiency of crack disease interpretation of radar data.

[0156] Based on the method provided by the present invention, the present invention also provides a three-dimensional ground penetrating radar multi-channel data splicing processing system, the system comprising:

[0157] A data acquisition and preprocessing module, used to acquire three-dimensional radar data of multiple measurement channels collected by a three-dimensional ground penetrating radar multi-channel antenna array and slice the three-dimensional radar data to obtain XY plane slice data of the three-dimensional radar data;

[0158] A data point longitude and latitude calculation module is used to calculate the longitude and latitude of the data points in the XY plane slice data to obtain the longitude and latitude coordinates of all data points;

[0159] An adjacent channel determination module, used to determine a reference channel among the multiple channels and to find all adjacent channels of the reference channel using an adjacent channel determination algorithm;

[0160] A drawing canvas range determination module, used to determine the canvas range using a canvas range determination algorithm according to the latitude and longitude coordinates of the data points in the reference survey track;

[0161] A data point range determination module, used to determine the data point ranges of the reference track and the adjacent track in the canvas respectively according to the canvas range;

[0162] The track data point stitching module is used to interpolate, translate and rotate the data points within the data point range, map the data points to pixel points, and generate a stitched image of the track horizontal plane.

[0163] The data point latitude and longitude calculation module specifically includes:

[0164] The data point longitude and latitude calculation unit is used to approximate the ground plane to the Euclidean plane, derive the longitude and latitude coordinates of all data points according to the longitude and latitude coordinates of the center of the multi-channel antenna array, the angle between the driving direction and the north direction, and the distance between other data points and the central measuring point in the XY plane slice data, and project all data points onto the satellite map according to the longitude and latitude coordinates of all data points.

[0165] The adjacent measurement track judgment module specifically includes:

[0166] A coordinate maximum value point acquisition unit, used to acquire the coordinate maximum value points of the data points in the reference measurement track and the measurement track to be determined;

[0167] An area marking unit, used for marking a first area with the coordinate maximum point in the reference track as a diagonal point, and marking a second area with the coordinate maximum point in the to-be-determined track as a diagonal point;

[0168] a ratio calculation unit, used to calculate the ratio of the intersection of the first area and the second area to the union of the first area and the second area;

[0169] The adjacent channel determination unit is used to determine that the channel to be determined is an adjacent channel of the reference channel if the ratio is greater than a preset ratio threshold.

[0170] The drawing canvas range determination module specifically includes:

[0171] A coordinate conversion unit, used to convert the latitude and longitude coordinates of the data points in the reference survey track into UTM coordinates, and set the central axis of the canvas to be consistent with the central axis of the data points in the intermediate survey track;

[0172] A starting boundary point selection unit, used to select a starting boundary point and a reference data point of the canvas from the data points of the first channel in the reference measurement track;

[0173] An end boundary point determination unit, used to determine the UTM coordinates of the end boundary point of the canvas according to the UTM coordinates of the start boundary point and the reference data point;

[0174] A canvas range boundary line determination unit is used to determine the boundary line equation of the canvas range according to the UTM coordinates of the starting boundary point, the reference data point and the ending boundary point.

[0175] For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0176] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for splicing and processing multi-channel data of three-dimensional ground penetrating radar, characterized in that: include: Acquire three-dimensional radar data of multiple measurement channels collected by a three-dimensional ground penetrating radar multi-channel antenna array and perform slicing processing on the three-dimensional radar data to obtain XY plane slicing data of the three-dimensional radar data; Calculate the longitude and latitude of the data points in the XY plane slice data to obtain the longitude and latitude coordinates of all the data points; The latitude and longitude calculation of the data points in the XY plane slice data to obtain the latitude and longitude coordinates of all the data points specifically includes: Approximating the ground plane to the Euclidean plane, deriving the longitude and latitude coordinates of all data points according to the longitude and latitude coordinates of the center of the multi-channel antenna array, the angle between the driving direction and the true north direction, and the distances between other data points and the central measuring point in the XY plane slice data, and projecting all data points onto the satellite map according to the longitude and latitude coordinates of all data points; Assume that point A is the recorded longitude and latitude coordinates, point B is the longitude and latitude coordinates of the data point to be sought, and point B is located southeast of point A. Then the longitude and latitude coordinates between AB satisfy the following formula: Among them, lon1 and lat1 are the longitude and latitude coordinates of point A, lon2 and lat2 are the longitude and latitude coordinates of point B; γ is the complementary angle of θ; ||AB|| is the distance between point A and point B; R0 is the average radius of the earth; θ is the angle between the driving direction and the true north direction; Determine a reference track among the multiple tracks and use an adjacent track determination algorithm to find all adjacent tracks of the reference track; The step of using an adjacent measurement track judgment algorithm to find all adjacent measurement tracks of the reference measurement track specifically includes: Obtaining the coordinate maximum points of the data points in the reference track and the track to be determined; The first area is marked with the coordinate maximum point in the reference track as the diagonal point, and the second area is marked with the coordinate maximum point in the to-be-determined track as the diagonal point; Calculate the ratio of the intersection of the first area and the second area to the union of the first area and the second area; If the ratio is greater than a preset ratio threshold, determining that the to-be-determined measurement track is an adjacent measurement track of the reference measurement track; Determine the canvas range using a canvas range determination algorithm according to the latitude and longitude coordinates of the data points in the reference survey track; Determining the canvas range using a canvas range determination algorithm according to the latitude and longitude coordinates of the data points in the reference survey track specifically includes: Convert the latitude and longitude coordinates of the data points in the reference track into UTM coordinates, and set the center axis of the canvas to be consistent with the center axis of the data points in the middle track; Selecting the starting boundary point and reference data point of the canvas from the data points of the first channel in the reference track; Select the data point of channel 1 in the reference channel and use point p i1 As the starting boundary point of the canvas; then select another point p in the data points of channel 1 j1 As a reference data point; Determine the UTM coordinates of the end boundary point of the canvas according to the UTM coordinates of the start boundary point and the reference data point; Assume that p i1 The UTM coordinates are (x1, y1), p j1 The UTM coordinates are (x2, y2), then the end boundary point p of the canvas edge0 The UTM coordinates (x edge0 ,y edge0 ) is obtained using the following formula: Where L is the set length of the canvas; ||p i1 p j1 || is the starting boundary point p i1 With reference data point p j1 The distance between Determine a boundary line equation of the canvas range according to the UTM coordinates of the starting boundary point, the reference data point, and the ending boundary point; Determine the data point ranges of the reference track and the adjacent track in the canvas respectively according to the canvas range; After interpolation, coordinate translation and rotation are performed on the data points within the data point range, the data points are mapped to pixel points to generate a spliced ​​image of the horizontal plane of the survey track.

2. The method according to claim 1, characterized in that The determining, according to the canvas range, respectively the data point ranges of the reference track and the adjacent track in the canvas specifically includes: Determining the data point range of the reference track in the canvas according to the boundary line equation of the canvas range; the boundary line of the canvas range includes a starting boundary line and an ending boundary line; Sequentially detect the distance between each data point in the adjacent measurement track and the canvas boundary line, determine the data point with the shortest distance from the starting boundary line as the starting point, and determine the data point with the shortest distance from the ending boundary line as the ending point; The data point range of the adjacent measurement track in the canvas is determined according to the coordinate range of the starting point and the ending point.

3. The method according to claim 1, characterized in that After interpolating, translating and rotating the data points within the data point range, mapping the data points to pixel points to generate a spliced ​​image of the horizontal plane of the measurement track, specifically includes: Mapping the data points within the data point range onto the canvas; Use bilinear interpolation to interpolate the data points on the canvas; After interpolation, the UTM coordinates of the data points on the canvas are translated, and the midpoint of the canvas is translated to the origin of the coordinates; After translating the coordinates, rotate the coordinate system with the center of the canvas as the origin; The data points after the rotation of the coordinate system are mapped to the image pixels, and the images of the horizontal planes of the measurement tracks are output one by one.

4. A three-dimensional ground penetrating radar multi-channel data splicing processing system, characterized in that: include: A data acquisition and preprocessing module, used to acquire three-dimensional radar data of multiple measurement channels collected by a three-dimensional ground penetrating radar multi-channel antenna array and slice the three-dimensional radar data to obtain XY plane slice data of the three-dimensional radar data; A data point longitude and latitude calculation module is used to calculate the longitude and latitude of the data points in the XY plane slice data to obtain the longitude and latitude coordinates of all data points; The data point latitude and longitude calculation module specifically includes: A data point longitude and latitude calculation unit is used to approximate the ground plane to a Euclidean plane, derive the longitude and latitude coordinates of all data points according to the longitude and latitude coordinates of the center of the multi-channel antenna array, the angle between the driving direction and the true north direction, and the distances between other data points and the central measuring point in the XY plane slice data, and project all data points onto a satellite map according to the longitude and latitude coordinates of all data points; Assume that point A is the recorded longitude and latitude coordinates, point B is the longitude and latitude coordinates of the data point to be sought, and point B is located southeast of point A. Then the longitude and latitude coordinates between AB satisfy the following formula: Among them, lon1 and lat1 are the longitude and latitude coordinates of point A, lon2 and lat2 are the longitude and latitude coordinates of point B; γ is the complementary angle of θ; ||AB|| is the distance between point A and point B; R0 is the average radius of the earth; θ is the angle between the driving direction and the true north direction; An adjacent channel determination module, used to determine a reference channel among the multiple channels and to find all adjacent channels of the reference channel using an adjacent channel determination algorithm; The adjacent measurement track judgment module specifically includes: A coordinate maximum value point acquisition unit, used to acquire the coordinate maximum value points of the data points in the reference measurement track and the measurement track to be determined; An area marking unit, used for marking a first area with the coordinate maximum point in the reference track as a diagonal point, and marking a second area with the coordinate maximum point in the to-be-determined track as a diagonal point; a ratio calculation unit, used to calculate the ratio of the intersection of the first area and the second area to the union of the first area and the second area; An adjacent channel determination unit, configured to determine that the channel to be determined is an adjacent channel of the reference channel if the ratio is greater than a preset ratio threshold; A drawing canvas range determination module, used to determine the canvas range using a canvas range determination algorithm according to the latitude and longitude coordinates of the data points in the reference survey track; The drawing canvas range determination module specifically includes: A coordinate conversion unit, used to convert the latitude and longitude coordinates of the data points in the reference survey track into UTM coordinates, and set the central axis of the canvas to be consistent with the central axis of the data points in the intermediate survey track; A starting boundary point selection unit, used to select a starting boundary point and a reference data point of the canvas from the data points of the first channel in the reference measurement track; Select the data point of channel 1 in the reference channel and use point p i1 As the starting boundary point of the canvas; then select another point p in the data points of channel 1 j1 As a reference data point; An end boundary point determination unit, used to determine the UTM coordinates of the end boundary point of the canvas according to the UTM coordinates of the start boundary point and the reference data point; Assume that p i1 The UTM coordinates are (x1, y1), p j1 The UTM coordinates are (x2, y2), then the end boundary point p of the canvas edge0 The UTM coordinates (x edge0 ,y edge0 ) is obtained using the following formula: Where L is the set length of the canvas; ||p i1 p j1 || is the starting boundary point p i1 With reference data point p j1 The distance between A canvas range boundary line determination unit, configured to determine a boundary line equation of the canvas range according to the UTM coordinates of the starting boundary point, the reference data point, and the ending boundary point; A data point range determination module, used to determine the data point ranges of the reference track and the adjacent track in the canvas respectively according to the canvas range; The track data point stitching module is used to interpolate, coordinate translate and rotate the data points within the data point range, map the data points to pixel points, and generate a track horizontal plane stitched image.

Images