Road surface misalignment detection method and device based on precise three-dimensional

By acquiring three-dimensional contour data of the road surface through a precise three-dimensional detection method, calculating the contour reference surface and deviation, and automatically detecting road misalignments, the problem of large errors, low efficiency and expensive equipment in existing technologies is solved, and efficient and low-cost misalignment detection is achieved.

CN114511495BActive Publication Date: 2024-11-05WUHAN WUDA ZOYON SCI & TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202111572820.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-12-21
Publication Date
2024-11-05
Estimated Expiration
2041-12-21

AI Technical Summary

Technical Problem

Existing road misalignment detection technologies suffer from problems such as large measurement errors, low detection efficiency, and high equipment costs.

Method used

A pavement misalignment detection method based on precision 3D is adopted. By acquiring the 3D contour data of the pavement, calculating the contour reference surface and deviation, identifying suspected joint points, performing noise reduction processing, extending the joint and calculating misalignment information, and realizing automated detection.

Benefits of technology

It improves detection efficiency, reduces detection errors, and eliminates the need for expensive equipment, thus lowering detection costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114511495B_ABST
    Figure CN114511495B_ABST
Patent Text Reader

Abstract

The present invention provides a road misalignment detection method and device based on precise three-dimensional, the method comprising: obtaining road surface three-dimensional contour data, and obtaining a contour reference surface and a contour deviation between a measuring point in the road surface three-dimensional contour data and the contour reference surface based on the road surface three-dimensional contour data; obtaining a suspected joint point based on the contour deviation, and obtaining a suspected joint denoised binary image based on the suspected joint point; obtaining an original joint target image and a joint representative position based on the row direction projection feature of the target measuring point in the suspected joint denoised binary image; obtaining a target joint binary image through a joint extension operation based on the suspected joint denoised binary image, the original joint target image, and the joint representative position; obtaining road surface misalignment information based on the target joint binary image and the road surface three-dimensional contour data. The road surface misalignment detection method and device based on precise three-dimensional provided by the present invention can reduce measurement errors and improve detection efficiency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of road surface detection, and in particular to a method and device for detecting road surface misalignment based on precise three-dimensional detection. Background Art

[0002] Pavement misalignment refers to the elevation difference between two adjacent cement slabs at the transverse joint of a cement concrete pavement. Misalignment is an important factor affecting the smoothness, lifespan, and reflective cracks of cement concrete pavements after overlaying. It is also an important indicator that must be considered when evaluating the technical condition of cement concrete pavements, and when designing maintenance and overlaying. Therefore, how to accurately and efficiently detect and evaluate misalignment has become a matter of great concern to highway management and maintenance departments.

[0003] At present, the commonly used methods for detecting misalignment are manual method and automatic detection method. The manual method uses a ruler, vernier caliper or level, which is slow, low-precision and interferes with traffic; the automatic detection method uses the fact that the existence of misalignment will significantly affect the characteristics of the International Roughness Index (IRI), and includes laser section instruments, ultrasonic section instruments, etc. The section instruments are expensive and cannot accurately measure the misalignment (due to only a small amount of longitudinal profile, the measurement results are easily affected by cracks, peeling, etc.).

[0004] In summary, the existing misalignment detection technology has problems such as large measurement error, low detection efficiency, and expensive equipment cost. Summary of the invention

[0005] The present invention provides a road misalignment detection method and device based on precise three-dimensional, which are used to solve the defects of large measurement error, low detection efficiency and expensive equipment cost in the prior art, and achieve the purpose of reducing measurement error and improving detection efficiency.

[0006] The present invention provides a road misalignment detection method based on precise three-dimensional technology, comprising:

[0007] Acquire road surface three-dimensional profile data, and obtain a profile reference surface and a profile deviation between a measuring point in the road surface three-dimensional profile data and the profile reference surface based on the road surface three-dimensional profile data;

[0008] Obtaining suspected seam points based on the contour deviation, and obtaining a suspected seam denoised binary image based on the suspected seam points;

[0009] Based on the row-direction projection features of the target measurement points in the suspected seam denoising binary image, an original seam target image and a seam representative position are obtained;

[0010] Based on the suspected seam denoising binary image, the original seam target image, and the seam representative position, a target seam binary image is obtained by performing a seam extension operation;

[0011] Based on the target joint binary image and the road surface three-dimensional contour data, road surface misalignment information is obtained.

[0012] According to the road misalignment detection method provided by the present invention, obtaining a suspected joint denoising binary image based on the suspected joint points includes:

[0013] Obtaining an original suspected seam point binary image based on the suspected seam points;

[0014] Taking the connected area in the original suspected seam point binary map as a unit, denoising is performed according to the length of the connected area, the amplitude of the contour deviation, and the directional characteristics of the connected area to obtain the suspected seam denoised binary map.

[0015] According to the road misalignment detection method provided by the present invention, the original joint target map and the joint representative position are obtained based on the row direction projection features of the target measurement points in the suspected joint denoising binary map, including:

[0016] Projecting the target points in the suspected seam denoising binary image along the row direction to obtain suspected seam points in each row;

[0017] Based on the statistical features of the rows of suspected seam points, the original seam target map and the seam representative positions are obtained.

[0018] According to the road misalignment detection method provided by the present invention, the original joint target map and the joint representative position are obtained based on the statistical characteristics of the suspected joint points in each row, including:

[0019] When the number of suspected seam points in the current row is greater than a preset number value, determining the suspected seam points in the current row as seam seed points;

[0020] Based on the seam seed points, obtaining the original seam target image;

[0021] The sum of seams with a row spacing less than a preset row spacing value is taken as a target seam row, and the target seam rows are merged into the same seam, and the row average value of the target seam rows is taken as the representative position of the seam;

[0022] The seam row is the row where the seam seed point is located.

[0023] According to the road misalignment detection method provided by the present invention, the target joint binary image is obtained by performing a joint extension operation based on the suspected joint denoising binary image, the original joint target image, and the joint representative position, including:

[0024] In the first target range corresponding to the seam seed point, the measurement points in the suspected seam denoising binary image that belong to the same connected area as the seam seed point are used as supplementary seam points;

[0025] Based on the seam seed points and the supplementary seam points, an extended seam binary map is obtained;

[0026] Searching for seam points in the extended seam binary image within a second target range corresponding to the seam representative position, and performing linear fitting on the searched seam points to obtain a fitted seam;

[0027] In the third target range corresponding to the seam representative position, the fitting seam position in the column where the missing seam exists in the extended seam binary image is used as the extended seam target point;

[0028] A target seam binary image is obtained based on the seam seed points, the supplementary seam points and the extended seam target points.

[0029] According to the road surface misalignment detection method provided by the present invention, the contour reference surface and the contour deviation between the measuring point in the road surface three-dimensional contour data and the contour reference surface are obtained based on the road surface three-dimensional contour data, including:

[0030] In the case where there are invalid measuring points in the road surface three-dimensional profile data, the invalid measuring points are replaced based on valid measuring points near the invalid measuring points to obtain new road surface three-dimensional profile data;

[0031] Processing the new road surface three-dimensional profile data based on a filtering method or a frequency domain variation method to obtain the profile reference surface;

[0032] Subtracting the new road surface three-dimensional profile data from the profile reference surface to obtain the profile deviation;

[0033] The suspected seam point is a measurement point whose corresponding contour deviation is greater than the segmentation threshold.

[0034] According to the road surface misalignment detection method provided by the present invention, the road surface misalignment information is obtained based on the target joint binary image and the road surface three-dimensional profile data, including:

[0035] Based on the target joint binary image, the representative elevations of the road surface on both sides before and after the joint are obtained;

[0036] The road surface misalignment information is obtained based on the representative elevations of the road surfaces on both sides before and after the joint.

[0037] According to the road misalignment detection method provided by the present invention, obtaining the representative elevations of the road surfaces before and after the joint based on the target joint binary image includes:

[0038] Based on the target seam binary image, with the fitting seam as the center, searching for a seam start point and a seam end point within a fourth target range;

[0039] In the fifth target range corresponding to the joint starting point, the average elevation of the road surface adjacent to the fitting joint is used as the representative elevation of the front road surface, and a representative elevation set of the front road surface is obtained; the front road surface is the road surface in the vehicle's forward direction;

[0040] Within the sixth target range corresponding to the joint end point, the average elevation of the road surface adjacent to the fitting joint is used as the rear side road surface representative elevation, and a rear side road surface representative elevation set is obtained.

[0041] According to the road surface misalignment detection method provided by the present invention, the road surface misalignment information is obtained based on the representative elevations of the road surfaces on both sides before and after the joint, including:

[0042] Based on the representative elevations of the road surfaces on both sides before and after the joint, at least one of the full width misalignment value, the left and right wheel track misalignment values, the maximum misalignment value, the full width misalignment mean value, the full width misalignment median value and the full width weighted misalignment value is obtained;

[0043] The calculation method of the full-width misalignment value is as follows:

[0044] The absolute difference between the representative elevation of the front road surface and the representative elevation of the rear road surface is calculated column by column to obtain multiple columns of full-width misalignment values; the column direction corresponds to the width direction of the road surface, and the row direction is perpendicular to the column direction;

[0045] The calculation method of the left and right wheel track misalignment value is as follows:

[0046] Based on the correspondence between the column position of the road surface three-dimensional profile data and the position in the road surface width direction, the full width misalignment values ​​of the corresponding columns of the left wheel track band and the right wheel track band are respectively selected as the left and right wheel track band misalignment values;

[0047] The calculation method of the full width weighted misalignment value is as follows:

[0048] Based on the multiple columns of full-width misalignment values, the minimum distance between each column of full-width misalignment values ​​and the corresponding columns of the left wheel track band and the right wheel track band in the road width direction is calculated.

[0049] Based on the minimum distance in the road width direction and the distance between the left wheel track and the right wheel track in the road width direction, the weight corresponding to each column of full width misalignment value is calculated;

[0050] The full-width weighted misalignment value is calculated based on the multiple columns of full-width misalignment values ​​and the weights corresponding to the respective columns of full-width misalignment values.

[0051] The present invention also provides a road misalignment detection device based on precise three-dimensional detection, comprising:

[0052] A first calculation module is used to obtain road surface three-dimensional profile data, and obtain a profile reference surface and a profile deviation between a measuring point in the road surface three-dimensional profile data and the profile reference surface based on the road surface three-dimensional profile data;

[0053] A second calculation module, configured to obtain a suspected seam point based on the contour deviation, and obtain a suspected seam denoised binary image based on the suspected seam point;

[0054] A third calculation module is used to obtain an original seam target map and a seam representative position based on the row direction projection features of the target measurement points in the suspected seam denoising binary map;

[0055] A fourth calculation module, configured to obtain a target seam binary image by performing a seam extension operation based on the suspected seam denoising binary image, the original seam target image, and the seam representative position;

[0056] The fifth calculation module is used to obtain road surface misalignment information based on the target joint binary image and the road surface three-dimensional contour data.

[0057] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of any of the above-mentioned methods for detecting road misalignment are implemented.

[0058] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of any of the above-mentioned road misalignment detection methods are implemented.

[0059] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the steps of any of the above-mentioned road surface misalignment detection methods are implemented.

[0060] The road misalignment detection method and device based on precision three-dimensional provided by the present invention obtains the three-dimensional contour data of the road surface, calculates the contour reference surface and the contour deviation, and further determines the suspected joint points to obtain the target joint binary map, and finally obtains the road misalignment information. The entire process of calculating the road misalignment information is completed automatically, and does not require manual detection, nor does it require expensive equipment such as laser section meters and ultrasonic section meters to detect road misalignment. Therefore, the road misalignment detection method provided by the present invention can not only improve the detection efficiency of road misalignment and reduce the detection error of road misalignment, but also does not require the use of expensive section meters, thereby reducing the cost of road misalignment detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the present invention or 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 some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0062] Figure 1 This is one of the flow charts of the road misalignment detection method provided by the present invention;

[0063] Figure 2 The present invention provides a first original suspected seam point binary image;

[0064] Figure 3 The present invention provides Figure 2 The corresponding seam positioning result diagram;

[0065] Figure 4 is a second original suspected seam point binary image provided by the present invention;

[0066] Figure 5 The present invention provides Figure 4 The corresponding seam positioning result diagram;

[0067] Figure 6 It is a structural schematic diagram of the road surface misalignment detection device provided by the present invention;

[0068] Figure 7 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0069] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are 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.

[0070] Combine the following Figure 1-Figure 7 The present invention describes a road misalignment detection method and device based on precise three-dimensional (3D) detection.

[0071] like Figure 1 As shown, the present invention provides a road misalignment detection method, comprising:

[0072] Step 110: Acquire road surface three-dimensional profile data, and obtain a profile reference surface and a profile deviation between a measuring point in the road surface three-dimensional profile data and the profile reference surface based on the road surface three-dimensional profile data.

[0073] It can be understood that the three-dimensional contour data of the road surface can be three-dimensional contour data obtained by line scanning the road surface.

[0074] In this embodiment, a three-dimensional road surface data acquisition system composed of two line scanning three-dimensional measurement sensors can be used as input to obtain precise three-dimensional road surface profile data.

[0075] Among them, the three-dimensional contour data of the road surface has been processed by three-dimensional modeling; the line scanning three-dimensional measurement sensor is composed of a three-dimensional camera, a laser, a controller, and a posture measurement sensor. The laser in the line scanning three-dimensional measurement sensor projects a laser line along the width direction of the road, and the line scanning three-dimensional measurement unit can obtain the road cross-sectional elevation data in a single measurement; the sampling spacing of the line scanning three-dimensional measurement unit in the cross-sectional direction is 1-3mm, for example, 1mm, and the sampling spacing in the driving direction is 1-5mm, for example, 5mm, and the measurement width is 2000-4000mm, for example, 3600mm.

[0076] Furthermore, after acquiring the three-dimensional road surface profile data, the three-dimensional road surface profile data may be preprocessed, including measurement abnormal value processing, profile reference surface acquisition, and profile deviation acquisition.

[0077] Step 120: obtaining suspected seam points based on the contour deviation, and obtaining a suspected seam denoised binary image based on the suspected seam points.

[0078] Step 130: Based on the row-direction projection features of the target measurement points in the suspected seam denoised binary image, an original seam target image and seam representative positions are obtained.

[0079] It can be understood that the suspected seam point is a measurement point that is determined to be a possible seam point after preliminary judgment. The seam representative position is based on the distribution information of the row where the seam is located, and the seam rows with a row distance less than the preset row distance value are merged into the same seam, corresponding to a representative position.

[0080] Step 140 : Based on the suspected seam denoising binary image, the original seam target image, and the seam representative position, a target seam binary image is obtained by performing a seam extension operation.

[0081] It can be understood that after obtaining the original seam target map and the seam representative position, seam extension can also be performed based on the original seam target map and the seam representative position, and the seam extension includes connected area extension and missing area extension, and finally a target seam binary map is obtained.

[0082] Step 150: Obtain road surface misalignment information based on the target joint binary image and the road surface three-dimensional contour data.

[0083] It can be understood that, based on the target joint binary map, the road surface misalignment information can be obtained by first obtaining the representative elevations of the road surfaces before and after the joint based on the joint binary map, and then calculating the road surface misalignment value.

[0084] In some embodiments, obtaining a suspected seam denoising binary image based on the suspected seam points includes:

[0085] Obtaining an original suspected seam point binary image based on the suspected seam points;

[0086] Taking the connected area in the original suspected seam point binary map as a unit, denoising is performed according to the length of the connected area, the amplitude of the contour deviation, and the directional characteristics of the connected area to obtain the suspected seam denoised binary map.

[0087] It can be understood that, based on the suspected seam points, a binary image of suspected seam points is obtained, which can be: the connected areas of the original binary image of suspected seam points are first marked, and then denoising is performed based on the length of the connected area, the contour deviation amplitude, and the directional characteristics of the connected area in units of the connected area to obtain a denoised binary image of suspected seams.

[0088] In some embodiments, based on the row-direction projection features of the target measurement points in the suspected seam denoised binary image, obtaining the original seam target image and the seam representative position includes:

[0089] Projecting the target points in the suspected seam denoising binary image along the row direction to obtain suspected seam points in each row;

[0090] Based on the statistical features of the rows of suspected seam points, the original seam target map and the seam representative positions are obtained.

[0091] The statistical features of each row of suspected seam points may be the number of suspected seam points in each row and the ratio of the number of suspected seam points in each row to the total number of suspected seam points.

[0092] In some embodiments, obtaining the original seam target map and the seam representative position based on the statistical features of each row of suspected seam points includes:

[0093] When the number of suspected seam points in the current row is greater than a preset number value, determining the suspected seam points in the current row as seam seed points;

[0094] Based on the seam seed points, obtaining the original seam target image;

[0095] The sum of seams with a row spacing less than a preset row spacing value is taken as a target seam row, and the target seam rows are merged into the same seam, and the row average value of the target seam rows is taken as the representative position of the seam;

[0096] The seam row is the row where the seam seed point is located.

[0097] It can be understood that seam target positioning includes obtaining seam seed points and obtaining seam representative positions.

[0098] The method for obtaining the seam seed points is as follows: the suspected seam denoising binary image is projected along the row direction (road direction), and the number of suspected seam points in each row is counted. If the number of suspected seam points in the current row is greater than the preset number value, the suspected seam points in the current row are marked as seam seed points, and the current row is recorded as the row where the seam is located. The binary image composed of the seam seed points is recorded as the original seam target image.

[0099] The method for obtaining the representative position of the seam is as follows: according to the distribution information of the row where the seam is located, the seam rows whose row distance is less than the preset row distance value are merged into the same seam, and the row mean is recorded as the representative position of the merged seam.

[0100] In some embodiments, the obtaining of the target seam binary image by performing a seam extension operation based on the suspected seam denoising binary image, the original seam target image, and the seam representative position includes:

[0101] In the first target range corresponding to the seam seed point, the measurement points in the suspected seam denoising binary image that belong to the same connected area as the seam seed point are used as supplementary seam points;

[0102] Based on the seam seed points and the supplementary seam points, an extended seam binary map is obtained;

[0103] Searching for seam points in the extended seam binary image within a second target range corresponding to the seam representative position, and performing linear fitting on the searched seam points to obtain a fitted seam;

[0104] In the third target range corresponding to the seam representative position, the fitting seam position in the column where the missing seam exists in the extended seam binary image is used as the extended seam target point;

[0105] A target seam binary image is obtained based on the seam seed points, the supplementary seam points and the extended seam target points.

[0106] It can be understood that the seam extension includes the extension of the connected area and the extension of the missing area.

[0107] The specific method for extending the connected area is: first, based on the seam seed points in the original seam target map, for each seam seed point, within the first target range, the points in the denoised suspected seam binary map that belong to the same connected area as the current seam seed point are supplemented as seam points to obtain an extended seam binary map.

[0108] The specific method of extending the missing area is as follows: based on the seam representative position information, for each seam, with the seam representative position as the center, find the seam point in the extended seam binary map within the second target range, and perform linear fitting on it to obtain the fitted seam. If there is a column of missing seams in the extended seam binary map within the second target range, the position of the fitted seam in the corresponding column is marked as the seam target point to obtain the target seam binary map.

[0109] In some embodiments, obtaining a profile reference surface and a profile deviation between a measuring point in the three-dimensional profile data of the road surface and the profile reference surface based on the three-dimensional profile data of the road surface includes:

[0110] In the case where there are invalid measuring points in the road surface three-dimensional profile data, the invalid measuring points are replaced based on valid measuring points near the invalid measuring points to obtain new road surface three-dimensional profile data;

[0111] Processing the new road surface three-dimensional profile data based on a filtering method or a frequency domain variation method to obtain the profile reference surface;

[0112] Subtracting the new road surface three-dimensional profile data from the profile reference surface to obtain the profile deviation;

[0113] The suspected seam point is a measurement point whose corresponding contour deviation is greater than the segmentation threshold.

[0114] It can be understood that after obtaining the contour reference surface, the measurement outliers of the contour reference surface can be processed, and the three-dimensional contour data of the road surface after the measurement outliers are processed can be filtered (such as median filtering, low-pass filtering, mean filtering, total variation filtering), or frequency domain transformation (such as Fourier transform, wavelet transform) to obtain the contour reference surface.

[0115] The contour deviation is obtained by measuring the difference between the three-dimensional contour data after outlier processing and the contour reference surface.

[0116] For each measuring point, the segmentation threshold is adaptively calculated according to the contour deviation near the measuring point. If the contour deviation value of the measuring point is greater than or equal to the segmentation threshold, the current measuring point is marked as a suspected seam point, and the original suspected seam point binary image is obtained.

[0117] In some embodiments, obtaining the road surface misalignment information based on the target joint binary image and the road surface three-dimensional contour data includes:

[0118] Based on the target joint binary image, the representative elevations of the road surface on both sides before and after the joint are obtained;

[0119] The road surface misalignment information is obtained based on the representative elevations of the road surfaces on both sides before and after the joint.

[0120] Further, obtaining the representative elevations of the road surfaces on both sides of the joint before and after the joint based on the target joint binary image includes:

[0121] Based on the target seam binary image, with the fitting seam as the center, searching for a seam start point and a seam end point within a fourth target range;

[0122] In the fifth target range corresponding to the starting point of the joint, the average elevation of the road surface adjacent to the fitting joint is used as the representative elevation of the front road surface, and a representative elevation set of the front road surface is obtained; the front road surface is the road surface in the forward direction of the vehicle, and the rear road surface is the road surface in the opposite direction of the forward direction of the vehicle;

[0123] Within the sixth target range corresponding to the joint end point, the average elevation of the road surface adjacent to the fitting joint is used as the rear side road surface representative elevation, and a rear side road surface representative elevation set is obtained.

[0124] It can be understood that the representative elevations of the road surfaces on both sides of the joint are the representative elevations of the road surfaces on both sides of the joint. Specifically, the representative elevations of the road surfaces on both sides of the joint include a front side road surface representative elevation set and a rear side road surface representative elevation set.

[0125] In some embodiments, obtaining the road surface misalignment information based on the representative elevations of the road surfaces on both sides before and after the joint includes:

[0126] Based on the representative elevations of the road surface on both sides before and after the joint, at least one of the full-width misalignment value, left and right wheel track misalignment value, maximum misalignment value, full-width misalignment mean value, full-width misalignment median value and full-width weighted misalignment value is obtained.

[0127] It can be understood that the road misalignment information includes at least one of the full-width misalignment value, the left and right wheel track band misalignment value, the maximum misalignment value, the full-width misalignment mean value, the full-width misalignment median value and the full-width weighted misalignment value.

[0128] The calculation method of the full-width misalignment value is as follows:

[0129] The absolute difference between the representative elevation of the front road surface and the representative elevation of the rear road surface is calculated column by column to obtain multiple columns of full-width misalignment values; the column direction corresponds to the width direction of the road surface, and the row direction is perpendicular to the column direction;

[0130] The calculation method of the left and right wheel track misalignment value is as follows:

[0131] Based on the correspondence between the column position of the road surface three-dimensional profile data and the position in the road surface width direction, the full width misalignment values ​​of the corresponding columns of the left wheel track band and the right wheel track band are respectively selected as the left and right wheel track band misalignment values;

[0132] The calculation method of the full width weighted misalignment value is as follows:

[0133] Based on the multiple columns of full-width misalignment values, the minimum distance between each column of full-width misalignment values ​​and the corresponding columns of the left wheel track band and the right wheel track band in the road width direction is calculated.

[0134] Based on the minimum distance in the road width direction and the distance between the left wheel track and the right wheel track in the road width direction, the weight corresponding to each column of full width misalignment value is calculated;

[0135] The full-width weighted misalignment value is calculated based on the multiple columns of full-width misalignment values ​​and the weights corresponding to the respective columns of full-width misalignment values.

[0136] It can be understood that the representative elevations of the road surfaces on both sides of the front and rear of the joint are obtained, that is, for each joint, the representative elevations of the road surfaces on both sides of the front and rear of the joint are obtained column by column. The specific method is as follows: for each column of the joint, the position of the fitted joint is taken as the center, and the starting point and the end point of the joint are found along the driving direction within the third target range; above the starting point of the joint, the average elevation of the road surface adjacent to the joint within the fourth target range is taken as the representative elevation of the front side road surface, and the representative elevation set of the front side road surface {FZ1, FZ2, …, FZ n}; Below the end point of the joint, take the average elevation of the road surface adjacent to the joint within the fifth target range as the representative elevation of the rear road surface {BZ1, BZ2,…, BZ n}, where n is the number of columns of the target seam binary image.

[0137] Misalignment calculation: that is, the misalignment value is calculated for each joint. The calculation methods include full width misalignment calculation, left and right wheel track misalignment calculation, maximum misalignment calculation, full width misalignment average calculation, full width misalignment median calculation, and full width weighted misalignment calculation.

[0138] Calculation of full width misalignment: Calculate the absolute difference between the representative elevation of the front road surface and the representative elevation of the rear road surface column by column, recorded as SS j ,{SS j |SS j =abs(FZ j -BZ j ),j=1,2,…,n}.

[0139] Calculation of the misalignment of the left and right wheel tracks: Based on the correspondence between the column position of the three-dimensional road surface data and the position in the road width direction, the full width misalignment corresponding to the corresponding column position of the left and right wheel tracks is selected as the misalignment of the left and right wheel tracks, which are recorded as S Left (S Left ={SS j |j=L}), S Right (S Right ={SS j|j=R}), where L and R are the column numbers corresponding to the left and right wheel track positions respectively.

[0140] Calculation of the maximum misalignment value: the maximum value of the full-width misalignment value, denoted as S Max , S Max =max{SS1,SS2,…,SS n}.

[0141] The average value of the full width misalignment is calculated as follows: The average value of the full width misalignment is denoted as S Avg ,

[0142] The median value of the full width misalignment is calculated as follows: The median value of the full width misalignment is denoted by S Mid .

[0143] The full width weighted misalignment value is calculated as the full width misalignment value set {SS j |SS j =abs(FZ j -BZ j ),j=1,2,…,n}, the specific calculation steps are as follows:

[0144] Adaptive weight calculation. Based on the minimum distance DIS between each column in the full width misalignment value and the columns where the left wheel track and the right wheel track are located in the width direction i The distance between the left and right wheel tracks in the width direction, calculate its own weight w j , the calculation formula is as follows:

[0145]

[0146] Among them, DIS j =min(|x j -x L |,|x j -x R |), j = 1, 2, ..., n, x j is the distance position of the jth column in the width direction, x L 、x R They are the distance positions of the left and right wheel tracks in the width direction respectively;

[0147] The full-scale weighted error value is calculated based on all the error values ​​in the error value set and the corresponding weights w i , calculate the full-scale weighted misalignment value S weight .

[0148]

[0149] The acquisition of misalignment attributes includes the acquisition of misalignment mileage position information and the acquisition of misalignment severity. The acquisition of misalignment mileage position information uses the measured mileage information of the calculation unit and the misalignment relative position information inside the calculation unit to calculate the misalignment mileage information; the acquisition of misalignment severity classifies the misalignment into multiple levels based on the size of the misalignment value and customer application requirements.

[0150] The three-dimensional road profile data, combined with specific data processing methods, can also be used to detect indicators such as road cracks and road rutting.

[0151] In other embodiments, the road surface misalignment detection method includes: data preprocessing, joint positioning, and misalignment information acquisition.

[0152] Data preprocessing, including measurement outlier processing, contour reference surface acquisition, and contour deviation acquisition.

[0153] Seam positioning, including suspected seam point segmentation, suspected seam point coarse denoising, seam target positioning, and seam extension.

[0154] Obtaining misalignment information, including obtaining representative elevations of the road surface on both sides before and after the joint, calculating misalignment values, and obtaining misalignment attributes.

[0155] Data preprocessing, including measurement outlier processing, contour reference surface acquisition, contour deviation acquisition, specifically:

[0156] Processing of measurement outliers: Use valid measurement points near the zero-value measurement point (i.e., invalid measurement point) to replace the zero-value measurement point.

[0157] The contour reference surface is obtained, and the three-dimensional contour data after measuring abnormal values ​​is processed is obtained through the mean filtering method.

[0158] The contour deviation is obtained by measuring the difference between the three-dimensional contour data after outlier processing and the contour reference surface.

[0159] Suspected seam point segmentation: For each measuring point, the segmentation threshold T1 (T1=2*Tex1) is adaptively calculated based on the mean value Tex1 of the contour deviation near the measuring point. If the contour deviation value of the measuring point is greater than or equal to the segmentation threshold T1, the current point is marked as a suspected seam point, and the original suspected seam point binary image is obtained.

[0160] For the rough denoising of suspected seam points, the connected areas of the original binary image of suspected seam points are first marked, and then denoising is performed based on the length thresholds LenT1 (LenT1=40mm) and LenT2 (LenT2=400mm) of the connected areas, the contour deviation threshold T2 (T2=1.3*Tex2), and the horizontal angle threshold AngleT (AngleT=30°) of the connected areas to obtain the denoised binary image of suspected seams.

[0161] Where Tex2 is the mean value of the contour deviation corresponding to all points in the current suspected seam point binary image.

[0162] Seam target positioning, including seam seed point acquisition and seam representative position acquisition.

[0163] The method for obtaining the seam seed points is as follows: project the suspected seam denoising binary image along the row direction (road direction), and count the number of suspected seam points in each row. If the number of suspected seam points in the current row is greater than T2 (T2 = n*0.2), the suspected seam points in the current row are marked as seam seed points, and the current row is recorded as the row where the seam is located. The binary image composed of the seam seed points is recorded as the original seam target image.

[0164] The method for obtaining the representative position of the joint is as follows: according to the distribution information of the row where the joint is located, the joint rows with a row distance less than T3 (T3 = 500 mm) are merged into the same joint, and the row mean is recorded as the representative position of the merged joint;

[0165] Seam extension, including connected area extension and missing area extension.

[0166] The specific method for extending the connected area is as follows: first, based on the seam seed points in the original seam target map, for each seam seed point, within the first target range (50 mm), the points in the suspected seam denoising binary map that belong to the same connected area as the current seam seed point are supplemented as seam points to obtain an extended seam binary map.

[0167] The specific method for extending the missing area is to find the seam point in the extended seam binary map within the second target range (300mm) based on the representative position information of the seam for each seam, with the seam representative position as the center, and perform linear fitting on it to obtain the fitted seam. If there is a column of missing seams in the extended seam binary map within a given range (500mm in row direction), the position of the fitted seam in the corresponding column is marked as the seam target point to obtain the target seam binary map.

[0168] Obtaining misalignment information, including obtaining representative elevations of the road surface on both sides before and after the joint, calculating misalignment values, and obtaining misalignment attributes.

[0169] The representative elevations of the road surface on both sides of the front and rear of the joint are obtained. For each joint, the representative elevations of the road surface on both sides of the front and rear of the joint are obtained column by column. The specific method is as follows: for each column of the joint, the position of the fitted joint is taken as the center, and the starting point and end point of the joint are found along the driving direction within the third target range (500mm in the row direction); above the starting point of the joint, the average elevation of the road surface adjacent to the joint within the fourth target range (50mm in the row direction) is taken as the representative elevation of the front side road surface, and the representative elevation set of the front side road surface {FZ1, FZ2, …, FZ n}; Below the end point of the joint, take the average elevation of the road surface adjacent to the joint within the fifth target range (50 mm in the row direction) as the representative elevation of the rear road surface {BZ1, BZ2,…, BZ n}, where n is the number of columns of the binary image (n=3600).

[0170] Misalignment calculation: calculate the misalignment value for each joint. The calculation methods include full width misalignment calculation, left and right wheel track misalignment calculation, maximum misalignment calculation, full width misalignment average calculation, full width misalignment median calculation, and full width weighted misalignment calculation.

[0171] The full width misalignment value is calculated by calculating the absolute difference between the representative elevation of the front road surface and the representative elevation of the rear road surface column by column, recorded as SS j ,{SS j |SS j =abs(FZ j -BZ j ),j=1,2,…,n}.

[0172] The calculation of the misalignment value of the left and right wheel track belts is based on the correspondence between the column position of the three-dimensional road surface data and the position in the road width direction. The full width misalignment value corresponding to the corresponding column position of the left and right wheel track belts is selected as the misalignment value of the left and right wheel track belts, which are recorded as S Left (S Left ={SS j |j=L}), S Right (S Right ={SS j |j=R}), where L and R are the column numbers corresponding to the left and right wheel track positions respectively (L=1100, R=2700).

[0173] The maximum misalignment value is calculated as the maximum value of the full-width misalignment value, S Max , S Max =max{SS1,SS2,…,SS n}.

[0174] The average value of the full width misalignment is calculated as the average value of the full width misalignment, denoted as S Avg ,

[0175] The median value of the full width misalignment is calculated as the median value of the full width misalignment, denoted as S Mid .

[0176] The full width weighted misalignment value is calculated as the full width misalignment value set {SS j |SS j =abs(FZ j -BZ j),j=1,2,…,n}, the specific calculation steps are as follows:

[0177] Adaptive weight calculation. Based on the minimum distance DIS between each column in the full width misalignment value and the columns where the left wheel track and the right wheel track are located in the width direction i The distance between the left and right wheel tracks in the width direction, calculate its own weight w j , the calculation formula is as follows:

[0178]

[0179] Among them, DIS j =min(|x j -x L |,|x j -x R |), j = 1, 2, ..., n, x j is the distance position of the jth column in the width direction, x L 、x R They are the distance positions of the left and right wheel tracks in the width direction respectively;

[0180] The full-scale weighted error value is calculated based on all the error values ​​in the error value set and the corresponding weights w i , calculate the full-scale weighted misalignment value S weight .

[0181]

[0182] The misalignment attribute acquisition includes misalignment mileage position information acquisition and misalignment severity acquisition. The misalignment mileage position information acquisition uses the measurement mileage information of the calculation unit and the misalignment relative position information inside the calculation unit to calculate the misalignment mileage information; the misalignment severity acquisition combines the misalignment value and customer application requirements to classify the misalignment into multiple levels. Specifically, misalignment values ​​<2.54mm are classified as mild misalignment, misalignment values ​​≥2.54mm and ≤5.08mm are classified as moderate misalignment, and misalignment values ​​≥5.08mm are classified as severe misalignment.

[0183] In the above embodiment, the first original suspected seam point binary image is as follows: Figure 2 As shown, the seam positioning result corresponding to the first original suspected seam point binary image is as follows Figure 3 shown.

[0184] The second original suspected seam point binary image is as follows Figure 4 As shown, the seam positioning result corresponding to the second original suspected seam point binary image is as follows Figure 5 shown.

[0185] Figure 2-3In the test, the average full-width misalignment of the joint is 2.07mm, the median full-width misalignment is 2.02mm, the left wheel track misalignment is 1.84mm, the right wheel track misalignment is 2.65mm, and the maximum misalignment is 4.13mm.

[0186] Figure 4-5 In the middle, the average full-width misalignment of the upper seam is 3.21mm, the median full-width misalignment is 3.23mm, the left wheel track misalignment is 3.65mm, the right wheel track misalignment is 4.55mm, and the maximum misalignment is 7.27mm.

[0187] Figure 4-5 In the middle and lower seams, the average full-width misalignment is 1.33mm, the median full-width misalignment is 0.77mm, the left wheel track misalignment is 3.62mm, the right wheel track misalignment is 0.87mm, and the maximum misalignment is 14.07mm.

[0188] The precise three-dimensional road profile data can also be used to detect indicators such as road cracks and road rutting based on a specific data processing method.

[0189] In summary, the road misalignment detection method provided by the present invention includes: acquiring road surface three-dimensional contour data, and obtaining a contour reference surface and a contour deviation between a measuring point in the road surface three-dimensional contour data and the contour reference surface based on the road surface three-dimensional contour data; obtaining a suspected joint point based on the contour deviation, and obtaining a suspected joint denoised binary image, an original joint target image, and a joint representative position based on the suspected joint point; obtaining an original joint target image and a joint representative position based on the row direction projection features of the target measuring point in the suspected joint denoised binary image; obtaining a target joint binary image through a joint extension operation based on the suspected joint denoised binary image, the original joint target image, and the joint representative position;

[0190] Based on the target joint binary image and the road surface three-dimensional contour data, road surface misalignment information is obtained.

[0191] In the road surface misalignment detection method provided by the present invention, after obtaining the three-dimensional contour data of the road surface, the contour reference surface and the contour deviation are calculated, and then the suspected joint points are further determined to obtain the target joint binary map, and finally the road surface misalignment information is obtained. The entire process of calculating the road surface misalignment information is completed automatically, and no manual detection is required, nor is expensive equipment such as laser section meters and ultrasonic section meters required for road surface misalignment detection. Therefore, the road surface misalignment detection method provided by the present invention can not only improve the detection efficiency of road surface misalignment and reduce the detection error of road surface misalignment, but also does not require the use of expensive section meters, thereby reducing the cost of road surface misalignment detection.

[0192] The misalignment detection method provided by the present invention has high measurement accuracy and can effectively avoid the influence of road cracks, peeling, etc. on the detection results; the misalignment detection method provided by the present invention has low cost and can share measurement equipment with detection indicators such as road damage and road rutting.

[0193] The road surface misalignment detection device provided by the present invention is described below. The road surface misalignment detection device described below and the road surface misalignment detection method described above can be referenced to each other.

[0194] like Figure 6 As shown, the road misalignment detection device 600 provided by the present invention includes: a first calculation module 610, a second calculation module 620, a third calculation module 630, a fourth calculation module 640 and a fifth calculation module 650.

[0195] The first calculation module 610 is used to obtain road surface three-dimensional contour data, and obtain a contour reference surface and a contour deviation between a measuring point in the road surface three-dimensional contour data and the contour reference surface based on the road surface three-dimensional contour data.

[0196] The second calculation module 620 is used to obtain suspected seam points based on the contour deviation, and obtain a suspected seam denoised binary image based on the suspected seam points.

[0197] The third calculation module 630 is used to obtain an original seam target image and seam representative positions based on the row-direction projection features of the target measurement points in the suspected seam denoised binary image.

[0198] The fourth calculation module 640 is used to obtain a target seam binary image through a seam extension operation based on the suspected seam denoising binary image, the original seam target image, and the seam representative position.

[0199] The fifth calculation module 650 is used to obtain road surface misalignment information based on the target joint binary image and the road surface three-dimensional contour data.

[0200] In some embodiments, the second calculation module 620 includes: a first binary image generating unit and a binary image processing unit.

[0201] The first binary image generating unit is used to obtain an original suspected seam point binary image based on the suspected seam points.

[0202] The binary image processing unit is used to perform denoising based on the connected area in the original suspected seam point binary image, according to the length of the connected area, the amplitude of the contour deviation, and the directional characteristics of the connected area, so as to obtain the suspected seam denoised binary image.

[0203] In some embodiments, the third calculation module 630 includes: a suspected point generating unit and a suspected point processing unit.

[0204] The suspected point generating unit is used to project the target points in the suspected seam denoising binary image along the row direction to obtain suspected seam points in each row.

[0205] The suspected point processing unit is used to obtain the original seam target image and the seam representative position based on the statistical features of the rows of suspected seam points.

[0206] In some embodiments, the suspected point processing unit includes: a determination unit, a seam map generation unit, and a seam position calculation unit.

[0207] The determining unit is used to determine the suspected seam points in the current row as seam seed points when the number of suspected seam points in the current row is greater than a preset number value;

[0208] The seam map generating unit is used to obtain the original seam target map based on the seam seed points;

[0209] The seam position calculation unit is used to take the seams with a row spacing less than a preset row spacing value as target seam rows, merge the target seam rows into the same seam, and take the row average value of the target seam rows as the seam representative position;

[0210] The seam row is the row where the seam seed point is located.

[0211] In some embodiments, the fourth calculation module 640 includes: a supplementing unit, an expanding unit, a fitting unit, an extending unit, and a second binary image generating unit.

[0212] The supplementing unit is used to take, within the first target range corresponding to the seam seed point, the measurement points in the suspected seam denoising binary image that belong to the same connected area as the seam seed point as supplementary seam points;

[0213] The expansion unit is used to obtain an expanded seam binary map based on the seam seed points and the supplementary seam points;

[0214] The fitting unit is used to search for seam points in the extended seam binary image within a second target range corresponding to the seam representative position, and perform linear fitting on the searched seam points to obtain a fitted seam;

[0215] The extending unit is used to use the fitting seam position in the column with missing seams in the extended seam binary image as the extended seam target point within the third target range corresponding to the seam representative position;

[0216] The second binary image generating unit is used to obtain a target seam binary image based on the seam seed points, the supplementary seam points and the extended seam target points.

[0217] In some embodiments, the first calculation module 610 includes: a data acquisition unit, a data processing unit, and a deviation calculation unit.

[0218] The data acquisition unit is used for replacing invalid measuring points based on valid measuring points near the invalid measuring points to obtain new three-dimensional road surface contour data when there are invalid measuring points in the three-dimensional road surface contour data;

[0219] The data processing unit is used to process the new road surface three-dimensional profile data based on a filtering method or a frequency domain variation method to obtain the profile reference surface;

[0220] The deviation calculation unit is used to subtract the new road surface three-dimensional profile data from the profile reference surface to obtain the profile deviation.

[0221] The suspected seam point is a measurement point whose corresponding contour deviation is greater than the segmentation threshold.

[0222] In some embodiments, the fifth calculation module 650 includes: an elevation calculation unit and a misalignment calculation unit.

[0223] The elevation calculation unit is used to obtain the representative elevations of the road surface on both sides of the joint based on the target joint binary image;

[0224] The misalignment calculation unit is used to obtain the road surface misalignment information based on the representative elevations of the road surfaces on both sides before and after the joint.

[0225] In some embodiments, the elevation calculation unit includes: a search unit, a first elevation determination unit, and a second elevation determination unit.

[0226] The search unit is used to search for a seam start point and a seam end point within a fourth target range based on the target seam binary image and with the fitting seam as the center;

[0227] The first elevation determination unit is used to use the average elevation of the road surface adjacent to the fitting joint as the representative elevation of the front road surface within the fifth target range corresponding to the joint starting point, and obtain a representative elevation set of the front road surface; the front road surface is the road surface in the vehicle's forward direction;

[0228] The second elevation determination unit is used to use the average elevation of the road surface adjacent to the fitting joint as the rear side road surface representative elevation within the sixth target range corresponding to the joint end point, to obtain a rear side road surface representative elevation set.

[0229] In some embodiments, the misalignment calculation unit is further used to: obtain at least one of the full-width misalignment value, the left and right wheel track misalignment value, the maximum misalignment value, the full-width misalignment mean value, the full-width misalignment median value and the full-width weighted misalignment value based on the representative elevations of the road surface on both sides before and after the joint.

[0230] The calculation method of the full-width misalignment value is as follows:

[0231] The absolute difference between the representative elevation of the front road surface and the representative elevation of the rear road surface is calculated column by column to obtain multiple columns of full-width misalignment values; the column direction corresponds to the width direction of the road surface, and the row direction is perpendicular to the column direction;

[0232] The calculation method of the left and right wheel track misalignment value is as follows:

[0233] Based on the correspondence between the column position of the road surface three-dimensional profile data and the position in the road surface width direction, the full width misalignment values ​​of the corresponding columns of the left wheel track band and the right wheel track band are respectively selected as the left and right wheel track band misalignment values;

[0234] The calculation method of the full width weighted misalignment value is as follows:

[0235] Based on the multiple columns of full-width misalignment values, the minimum distance between each column of full-width misalignment values ​​and the corresponding columns of the left wheel track band and the right wheel track band in the road width direction is calculated.

[0236] Based on the minimum distance in the road width direction and the distance between the left wheel track and the right wheel track in the road width direction, the weight corresponding to each column of full width misalignment value is calculated;

[0237] The full-width weighted misalignment value is calculated based on the multiple columns of full-width misalignment values ​​and the weights corresponding to the respective columns of full-width misalignment values.

[0238] The first calculation module 610 is used to obtain road surface three-dimensional contour data, and obtain a contour reference surface and a contour deviation between a measuring point in the road surface three-dimensional contour data and the contour reference surface based on the road surface three-dimensional contour data.

[0239] The second calculation module 620 is used to obtain suspected seam points based on the contour deviation, and obtain an original seam target image and a seam representative position based on the suspected seam points.

[0240] The third calculation module 630 is used to obtain a target seam binary image based on the original seam target image and the seam representative position.

[0241] The fourth calculation module 640 is used to obtain road surface misalignment information based on the target joint binary map.

[0242] In some embodiments, the second calculation module 620 includes: a first binary image generating unit and a seam calculating unit.

[0243] The first binary image generating unit is used to obtain a suspected seam point binary image based on the suspected seam points.

[0244] The seam calculation unit is used to project the suspected seam point binary image along the row direction to obtain each row of suspected seam points, and obtain the original seam target image and the seam representative position based on the each row of suspected seam points.

[0245] In some embodiments, the seam calculation unit includes: a seed point determination unit, a seam target map calculation unit, and a seam representative position calculation unit.

[0246] The seed point determination unit is used to determine the suspected seam points in the current row as seam seed points when the number of suspected seam points in the current row is greater than a preset number value.

[0247] The seam target map calculation unit is used to obtain the original seam target map based on the seam seed points.

[0248] The seam representative position calculation unit is used to take the seams with a row spacing less than a preset row spacing value as target seam rows, merge the target seam rows into the same seam, and take the row average value of the target seam rows as the seam representative position.

[0249] The seam row is the row where the seam seed point is located.

[0250] In some embodiments, the third calculation module 630 includes: a supplementary seam point calculation unit, an expansion unit, a fitting unit, a seam target calculation unit, and a second binary image generation unit.

[0251] The supplementary seam point calculation unit is used to take, within the first target range corresponding to the seam seed point, the measurement points in the suspected seam point binary map that belong to the same connected area as the seam seed point as the supplementary seam points;

[0252] The expansion unit is used to obtain an expanded seam binary map based on the supplementary seam points;

[0253] The fitting unit is used to search for seam points in the extended seam binary image within a second target range corresponding to the seam representative position, and perform linear fitting on the searched seam points to obtain a fitted seam;

[0254] The seam target calculation unit is used to take the fitting seam position in the column with missing seams in the extended seam binary image as the seam target point within the third target range corresponding to the seam representative position;

[0255] The second binary image generating unit is used to obtain a target seam binary image based on the extended seam target point.

[0256] In some embodiments, the first calculation module 610 includes: a denoising unit, a reference surface calculation unit, and a deviation calculation unit.

[0257] The denoising unit is used for replacing invalid measuring points based on valid measuring points near the invalid measuring points to obtain new three-dimensional road surface profile data when there are invalid measuring points in the three-dimensional road surface profile data.

[0258] The reference surface calculation unit is used to process the new road surface three-dimensional profile data based on a filtering method or a frequency domain variation method to obtain the profile reference surface.

[0259] The deviation calculation unit is used to subtract the new road surface three-dimensional profile data from the profile reference surface to obtain the profile deviation.

[0260] The suspected seam point is a measurement point whose corresponding contour deviation is greater than the segmentation threshold.

[0261] In some embodiments, the fourth calculation module 640 is further used to obtain the full width misalignment value, left and right wheel track misalignment value, maximum misalignment value, full width misalignment mean value, full width misalignment median value and full width weighted misalignment value based on the target seam binary image.

[0262] The electronic device, computer program product and storage medium provided by the present invention are described below. The electronic device, computer program product and storage medium described below can be referenced to each other with the road misalignment detection method described above.

[0263] Figure 7 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 7 As shown, the electronic device may include: a processor 710, a communication interface 720, a memory 730 and a communication bus 740, wherein the processor 710, the communication interface 720 and the memory 730 communicate with each other through the communication bus 740. The processor 710 may call the logic instructions in the memory 730 to execute the road misalignment detection method, which includes:

[0264] Step 110, acquiring road surface three-dimensional profile data, and obtaining a profile reference surface based on the road surface three-dimensional profile data, and a profile deviation between a measuring point in the road surface three-dimensional profile data and the profile reference surface;

[0265] Step 120, obtaining suspected seam points based on the contour deviation, and obtaining the suspected seam denoising binary image, the original seam target image and the seam representative position based on the suspected seam points;

[0266] Step 130, obtaining an original seam target map and seam representative positions based on the row-direction projection features of the target measurement points in the suspected seam denoised binary map;

[0267] Step 140, obtaining a target seam binary image by performing a seam extension operation based on the suspected seam denoising binary image, the original seam target image, and the seam representative position;

[0268] Step 150: Obtain road surface misalignment information based on the target joint binary image and the road surface three-dimensional contour data.

[0269] In addition, the logic instructions in the above-mentioned memory 730 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0270] On the other hand, the present invention further provides a computer program product, the computer program product includes a computer program, the computer program can be stored in a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute the road misalignment detection method provided by the above methods, the method includes:

[0271] Step 110, acquiring road surface three-dimensional profile data, and obtaining a profile reference surface based on the road surface three-dimensional profile data, and a profile deviation between a measuring point in the road surface three-dimensional profile data and the profile reference surface;

[0272] Step 120, obtaining suspected seam points based on the contour deviation, and obtaining the suspected seam denoising binary image, the original seam target image and the seam representative position based on the suspected seam points;

[0273] Step 130, obtaining an original seam target map and seam representative positions based on the row-direction projection features of the target measurement points in the suspected seam denoised binary map;

[0274] Step 140, obtaining a target seam binary image by performing a seam extension operation based on the suspected seam denoising binary image, the original seam target image, and the seam representative position;

[0275] Step 150: Obtain road surface misalignment information based on the target joint binary image and the road surface three-dimensional contour data.

[0276] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the road misalignment detection method provided by the above methods is implemented, and the method comprises:

[0277] Step 110: Acquire road surface three-dimensional profile data, and obtain a profile reference surface based on the road surface three-dimensional profile data, and a profile deviation between a measuring point in the road surface three-dimensional profile data and the profile reference surface;

[0278] Step 120, obtaining suspected seam points based on the contour deviation, and obtaining the suspected seam denoising binary image, the original seam target image and the seam representative position based on the suspected seam points;

[0279] Step 130, obtaining an original seam target map and seam representative positions based on the row-direction projection features of the target measurement points in the suspected seam denoised binary map;

[0280] Step 140, obtaining a target seam binary image by performing a seam extension operation based on the suspected seam denoising binary image, the original seam target image, and the seam representative position;

[0281] Step 150: Obtain road surface misalignment information based on the target joint binary image and the road surface three-dimensional contour data.

[0282] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0283] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0284] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A road misalignment detection method based on precise three-dimensional, characterized in that: include: Acquire road surface three-dimensional profile data, and obtain a profile reference surface and a profile deviation between a measuring point in the road surface three-dimensional profile data and the profile reference surface based on the road surface three-dimensional profile data; Obtaining suspected seam points based on the contour deviation, and obtaining a suspected seam denoised binary image based on the suspected seam points; Based on the row-direction projection features of the target measurement points in the suspected seam denoising binary image, an original seam target image and a seam representative position are obtained; Based on the suspected seam denoising binary image, the original seam target image, and the seam representative position, a target seam binary image is obtained by performing a seam extension operation; Obtaining road surface misalignment information based on the target joint binary image and the road surface three-dimensional contour data; The step of obtaining the original seam target map and the seam representative position based on the row-direction projection features of the target measurement points in the suspected seam denoising binary map comprises: Projecting the target points in the suspected seam denoising binary image along the row direction to obtain suspected seam points in each row; Based on the statistical characteristics of the suspected seam points in each row, the original seam target map and the seam representative positions are obtained; The obtaining of the original seam target map and the seam representative position based on the statistical features of the suspected seam points in each row includes: When the number of suspected seam points in the current row is greater than a preset number value, determining the suspected seam points in the current row as seam seed points; Based on the seam seed points, obtaining the original seam target image; The sum of seams with a row spacing less than a preset row spacing value is taken as a target seam row, and the target seam rows are merged into the same seam, and the row average value of the target seam rows is taken as the representative position of the seam; Wherein, the seam row is the row where the seam seed point is located; The step of obtaining a target seam binary image by performing a seam extension operation based on the suspected seam denoising binary image, the original seam target image, and the seam representative position includes: In the first target range corresponding to the seam seed point, the measurement points in the suspected seam denoising binary image that belong to the same connected area as the seam seed point are used as supplementary seam points; Based on the seam seed points and the supplementary seam points, an extended seam binary map is obtained; Searching for seam points in the extended seam binary image within a second target range corresponding to the seam representative position, and performing linear fitting on the searched seam points to obtain a fitted seam; In the third target range corresponding to the seam representative position, the fitting seam position in the column where the missing seam exists in the extended seam binary image is used as the extended seam target point; A target seam binary image is obtained based on the seam seed points, the supplementary seam points and the extended seam target points.

2. The method for detecting road misalignment according to claim 1, characterized in that: The step of obtaining a suspected seam denoising binary image based on the suspected seam points includes: Obtaining an original suspected seam point binary image based on the suspected seam points; Taking the connected area in the original suspected seam point binary map as a unit, denoising is performed according to the length of the connected area, the amplitude of the contour deviation, and the directional characteristics of the connected area to obtain the suspected seam denoised binary map.

3. The method for detecting road misalignment based on precise three-dimensional detection according to claim 1, characterized in that: The step of obtaining a profile reference surface and a profile deviation between a measuring point in the three-dimensional profile data of the road surface and the profile reference surface based on the three-dimensional profile data of the road surface comprises: In the case where there are invalid measuring points in the road surface three-dimensional profile data, the invalid measuring points are replaced based on valid measuring points near the invalid measuring points to obtain new road surface three-dimensional profile data; Processing the new road surface three-dimensional profile data based on a filtering method or a frequency domain variation method to obtain the profile reference surface; Subtracting the new road surface three-dimensional profile data from the profile reference surface to obtain the profile deviation; The suspected seam point is a measurement point whose corresponding contour deviation is greater than the segmentation threshold.

4. The method for detecting road misalignment based on precise three-dimensional data according to any one of claims 1 to 3, characterized in that: The step of obtaining the road surface misalignment information based on the target joint binary image and the road surface three-dimensional profile data includes: Based on the target joint binary image, the representative elevations of the road surface on both sides before and after the joint are obtained; The road surface misalignment information is obtained based on the representative elevations of the road surfaces on both sides before and after the joint.

5. The method for detecting road misalignment based on precise three-dimensional detection according to claim 4 is characterized in that: The step of obtaining the representative elevations of the road surfaces on both sides of the joint before and after the joint based on the target joint binary image includes: Based on the target seam binary image, with the fitting seam as the center, searching for a seam start point and a seam end point within a fourth target range; In the fifth target range corresponding to the joint starting point, the average elevation of the road surface adjacent to the fitting joint is used as the representative elevation of the front road surface, and a representative elevation set of the front road surface is obtained; the front road surface is the road surface in the vehicle's forward direction; Within the sixth target range corresponding to the joint end point, the average elevation of the road surface adjacent to the fitting joint is used as the rear side road surface representative elevation, and a rear side road surface representative elevation set is obtained.

6. The method for detecting road misalignment based on precise three-dimensional detection according to claim 4 is characterized in that: The step of obtaining the road surface misalignment information based on the representative elevations of the road surfaces on both sides before and after the joint includes: Based on the representative elevations of the road surfaces on both sides before and after the joint, at least one of the full width misalignment value, the left and right wheel track misalignment values, the maximum misalignment value, the full width misalignment mean value, the full width misalignment median value and the full width weighted misalignment value is obtained; The calculation method of the full-width misalignment value is as follows: The absolute difference between the representative elevation of the front road surface and the representative elevation of the rear road surface is calculated column by column to obtain multiple columns of full-width misalignment values; the column direction corresponds to the width direction of the road surface, and the row direction is perpendicular to the column direction; The calculation method of the left and right wheel track misalignment value is as follows: Based on the correspondence between the column position of the road surface three-dimensional profile data and the position in the road surface width direction, the full width misalignment values ​​of the corresponding columns of the left wheel track band and the right wheel track band are respectively selected as the left and right wheel track band misalignment values; The calculation method of the full width weighted misalignment value is as follows: Based on the multiple columns of full-width misalignment values, the minimum distance between each column of full-width misalignment values ​​and the corresponding columns of the left wheel track band and the right wheel track band in the road width direction is calculated. Based on the minimum distance in the road width direction and the distance between the left wheel track and the right wheel track in the road width direction, the weight corresponding to each column of full width misalignment value is calculated; The full-width weighted misalignment value is calculated based on the multiple columns of full-width misalignment values ​​and the weights corresponding to the respective columns of full-width misalignment values.

7. A device for detecting road misalignment based on precise three-dimensional basis corresponding to the method for detecting road misalignment based on precise three-dimensional basis as claimed in claim 1, characterized in that: include: A first calculation module is used to obtain road surface three-dimensional profile data, and obtain a profile reference surface and a profile deviation between a measuring point in the road surface three-dimensional profile data and the profile reference surface based on the road surface three-dimensional profile data; A second calculation module, configured to obtain a suspected seam point based on the contour deviation, and obtain a suspected seam denoised binary image based on the suspected seam point; A third calculation module is used to obtain an original seam target map and a seam representative position based on the row direction projection features of the target measurement points in the suspected seam denoising binary map; A fourth calculation module, configured to obtain a target seam binary image by performing a seam extension operation based on the suspected seam denoising binary image, the original seam target image, and the seam representative position; The fifth calculation module is used to obtain road surface misalignment information based on the target joint binary image and the road surface three-dimensional contour data.