A method for analyzing the evolution law of road potholes

Abstract

The present invention discloses a method for analyzing the evolution law of road potholes, and belongs to the field of road detection. Based on the depth information and contour information of the potholes, the present invention constructs a pothole depth distribution factor and an expansion factor, which are respectively used to characterize the depth distribution and plane distribution of the potholes. By collecting the depth distribution factor and the expansion factor corresponding to different times, the change gradient of the pothole depth distribution factor and the change gradient of the expansion factor are calculated, and the analysis of the evolution law of the potholes in the depth and plane directions is realized. The development direction and development speed of the potholes can be accurately judged, and then the formation cause of the potholes can be inferred. The pothole diseases can be repaired by taking scientific maintenance measures, the maintenance cost can be reduced, and the safety of vehicle driving can be improved.

Description

Technical Field

[0001] The present invention belongs to the field of road detection, and more specifically, relates to a method for analyzing the evolution law of road potholes. Background Art

[0002] Roads are important infrastructure for vehicles and pedestrians to pass through. The service status of roads affects the travel effect of vehicles and pedestrians. The number of cars in my country has exceeded 280 million. The traditional rigid travel mode can no longer meet people's growing demand for transportation and travel for a better life. It is urgent to further improve people's travel quality from the perspective of road facilities. The repair of road surface diseases is an urgent problem to be solved to improve the quality of travel. Among them, potholes are one of the common diseases of road pavement, and potholes have a great impact on the safe driving of vehicles. When the tires of vehicles pass through potholes, there will be bumps. In severe cases, the driving direction will be changed, posing a serious threat to driving safety. At the same time, the reaction force on the vehicle at the pothole is large, which reduces the service life of tires and vehicles. Therefore, it is necessary to timely discover pothole diseases on the road surface and formulate reasonable maintenance measures to reduce the impact of potholes on people's travel. On the other hand, the repair effect of potholes depends on the accuracy of the analysis of the causes of pothole formation. Accurate analysis of the causes of diseases can ensure the rationality of pothole repair decisions, so as to achieve the purpose of reducing maintenance costs and improving road service quality.

[0003] At present, the detection of road potholes is usually done by taking photos with a camera and then identifying them based on the two-dimensional features of the pothole image. Since it is impossible to directly obtain the depth information of the pothole, it is difficult to directly obtain the volume information. With the development of detection technology, structured light scanning technology has become more mature and can realize the collection of three-dimensional information of the road surface in the horizontal direction and depth direction under driving conditions. On this basis, the volume of the pothole can be calculated by processing the collected data, and the calculation results can be used to guide maintenance decisions and the engineering quantity measurement of the disease repair materials.

[0004] However, current research and inventions have two problems when identifying and calculating three-dimensional potholes: (1) In terms of pothole identification efficiency, three-dimensional potholes contain massive data, and traditional manual identification methods are inefficient, while intelligent identification methods require high computer computing power and consume a lot of resources; (2) In the evaluation index of the evolution law of potholes on the road surface, only the volume value is usually used to reflect the changes in potholes, which cannot characterize the diversity of the evolution law of potholes and the development direction and speed of potholes. The mastery of the development direction and speed of potholes is the basis for accurately determining the causes of potholes, and then accurately formulating maintenance strategies to effectively ensure driving safety.

[0005] In view of the above problems, the present invention proposes a method for analyzing the evolution law of road potholes. Summary of the invention

[0006] In view of the above defects or improvement needs of the prior art, the present invention provides a method for analyzing the evolution laws of road potholes, which aims to solve the technical problems that the existing methods are unable to characterize the diversity of pothole evolution laws and the development direction and speed of potholes.

[0007] To achieve the above object, the present invention provides a method for analyzing the evolution law of road potholes, comprising:

[0008] S1. Collecting three-dimensional point cloud information of the road surface and marking the non-pothole area and the pothole area to obtain a three-dimensional matrix L; converting the three-dimensional matrix L into a two-dimensional matrix M containing pothole depth information by plane projection;

[0009] S2. Sort the elements in the two-dimensional matrix M according to the depth, count the number of elements corresponding to different depth values, and calculate the proportion of the number of elements corresponding to different depth values ​​relative to the total number of elements;

[0010] S3. according to the difference in users' attention to different pit depths, weights are assigned to the proportions corresponding to different depth values, and a pit depth distribution factor is calculated; the pit depth distribution factor represents the development of the pit along the depth direction;

[0011] S4. Calculate the average area of ​​the edge matrix corresponding to the entire pit to obtain a pit expansion factor; the pit expansion factor represents the development of the pit along the plane direction;

[0012] S5. sequentially collect the pit 3D point cloud information corresponding to different time periods, and execute steps S1-S4 for the pit 3D point cloud information collected at each time period to obtain the pit distribution factor and pit expansion factor at the corresponding time period;

[0013] S6. Based on the change gradient of the pit distribution factor and the change gradient of the pit expansion factor, the pit evolution law is obtained.

[0014] Furthermore, the pit depth distribution factor is:

[0015]

[0016] Nhi represents the number of different depth values, S represents the total number of elements in the two-dimensional matrix, q represents the weight, and h represents the i Represents the elements corresponding to different depth values.

[0017] Furthermore, the two-dimensional matrix M is constructed as follows:

[0018]

[0019] x and y are coordinates in two directions respectively, z is the vertical axis direction of the matrix L, with vertical downward as the positive direction; N is the number of pages of the three-dimensional matrix L, i is the i-th page, i=1,2,3,…,N, N represents the number of pages of the three-dimensional matrix.

[0020] Furthermore, step S4 specifically includes:

[0021] Sequentially extract the i-th page L(x, y, i) of the three-dimensional matrix L to obtain the matrix Li, and extract the edge contour matrix Fi of Li;

[0022] Perform the following operations on matrix Li, matrix B and matrix C respectively:

[0023] Li B =Li[x,y]*B[x,y]

[0024] Li C =Li[x,y]*C[x,y]

[0025] In the formula, * represents the convolution operation, Li B It is the new matrix formed by the operation of the image matrix Li and the matrix B; Li C It is a new matrix formed by the operation of the image matrix Li and the matrix C; the matrix B and the matrix C represent the edge operator;

[0026] Calculate Li B and Li C The maximum value at the same position in the matrix is ​​used to obtain the pit edge matrix Fi;

[0027] Calculate the area A of the pit edge matrix Fi Fi =∑Fi(x,y);

[0028] Calculate the average area Z of the edge matrix corresponding to the entire pit and get the pit expansion factor

[0029] Furthermore, before executing step S6, the method further includes correcting the pit depth distribution factor and the expansion factor according to the system error to obtain a pit correction depth distribution factor U' q and the pit correction expansion factor Z'.

[0030] Furthermore, the pit correction depth distribution factor U' q And the pit correction expansion factor Z' is:

[0031]

[0032]

[0033] V represents the systematic error.

[0034] Further, step S1 includes:

[0035] The three-dimensional point cloud information of the road surface is converted into a three-dimensional matrix H; and a pothole edge matrix is ​​established for each layer of the three-dimensional point cloud information of the road surface;

[0036] The three-dimensional matrix H is cut by using the matrix formed by the non-zero values ​​in the pit edge matrix to form a three-dimensional matrix J representing the circumscribed cuboid of the three-dimensional pit region;

[0037] The plane K parallel to the road surface is translated from top to bottom or from bottom to top to intersect with the matrix J. During the intersection process, the non-pothole area and the pothole area are marked respectively to obtain the three-dimensional matrix L.

[0038] In general, compared with the prior art, the above technical solution conceived by the present invention can achieve the following beneficial effects.

[0039] (1) Potholes are formed due to various factors, including vehicle speed, vehicle load, road material, environment, etc. Potholes formed due to different causes have different development characteristics and corresponding pothole repair measures. However, traditional 3D detection methods cannot analyze the development direction of potholes and it is difficult to propose scientific maintenance measures for pothole repair, resulting in poor pothole repair effects, repeated occurrence of diseases, and increased maintenance costs.

[0040] Based on the depth information and contour information of the potholes, the present invention constructs a pothole depth distribution factor and an expansion factor, which are respectively used to characterize the depth distribution and plane distribution of the potholes. By collecting the depth distribution factor and the expansion factor corresponding to different times, the change gradient of the pothole depth distribution factor and the change gradient of the expansion factor are calculated, and the analysis of the evolution law of the potholes in both depth and plane directions is realized. The development direction and speed of the potholes can be accurately judged, and the causes of the formation of the potholes can be inferred. Scientific maintenance measures can be taken to repair pothole diseases, reduce maintenance costs, and improve vehicle driving safety.

[0041] (2) The plane projection method proposed in the present invention can project the position information onto a plane, thereby achieving the purpose of reducing the dimension of the pit matrix. At the same time, the point cloud depth information is retained in the form of plane matrix elements on the two-dimensional plane matrix, thereby achieving efficient conversion of the three-dimensional pit matrix into a two-dimensional matrix, solving the problem of high computer computing power requirements and long calculation time for three-dimensional data processing; at the same time, without increasing the system computing load, the information on the pit depth direction is taken into account, thereby improving the accuracy of the analysis.

[0042] (3) The present invention adopts a mathematical statistics method to obtain the distribution frequency and proportion of different pit depth values, and introduces a weight parameter, which can be used to characterize the weight difference of different pit depths and reflect the degree of attention paid by maintenance managers to the pit depth.

[0043] (4) The present invention takes into account the influence of system errors when calculating the depth distribution factor and the expansion factor, thereby improving the accuracy of the calculation results. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 The present invention is a flow chart of a method for analyzing the evolution law of road potholes. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0046] Combination Figure 1 , the method of the present invention comprises the following steps:

[0047] S1. Collecting three-dimensional point cloud information of the road surface and marking the non-pothole area and the pothole area to obtain a three-dimensional matrix L; converting the three-dimensional matrix L into a two-dimensional matrix M containing pothole depth information by plane projection;

[0048] Step S1 specifically includes the following process:

[0049] Collect the three-dimensional point cloud information of the road surface and convert it into a three-dimensional matrix H; and establish a pothole edge matrix for each layer of the three-dimensional point cloud information of the road surface;

[0050] The three-dimensional matrix H is cut by using the matrix formed by the non-zero values ​​in the pit edge matrix to form a three-dimensional matrix J representing the circumscribed rectangular parallelepiped of the three-dimensional pit region; specifically, the non-zero values ​​in the matrix F are searched to form a new matrix G; the three-dimensional matrix H is cut by using the matrix G to form a three-dimensional matrix J, which is the circumscribed rectangular parallelepiped matrix of the three-dimensional pit region;

[0051] The plane K parallel to the road surface is translated from top to bottom or from bottom to top to intersect with the matrix J. During the intersection process, the non-pothole area and the pothole area are marked respectively to obtain a three-dimensional matrix L. Specifically, the translation step length is 1 pixel. During the intersection process between the plane K and the three-dimensional matrix J, the non-pothole area is marked as 0 and the pothole area is marked as 1, thereby obtaining a three-dimensional matrix L consisting of only 0 and 1 in the three-dimensional matrix J.

[0052] The three-dimensional matrix L is transformed into a two-dimensional matrix M containing the pit depth information by plane projection;

[0053] The above steps establish a three-dimensional matrix of the pits. Since the three-dimensional matrix calculation process requires high computer computing power, occupies a large number of resources, and has high calculation costs, the present invention reduces the resource requirements for matrix calculation by converting the three-dimensional matrix of the pits into a two-dimensional matrix.

[0054] The existing plane projection method mainly focuses on the projection of point cloud position information. After the projection is completed, the information in the depth direction of the point cloud is deleted, and only the plane information can be analyzed. In view of this situation, the present invention improves the plane projection method. The plane projection method proposed in this invention can project the position information onto the plane, thereby achieving the purpose of reducing the dimension of the pit matrix. At the same time, the point cloud depth information is retained in the form of plane matrix elements on the two-dimensional plane matrix.

[0055] Based on the three-dimensional matrix L, a new two-dimensional matrix M is constructed as follows:

[0056]

[0057] Wherein, z is the vertical axis direction of the matrix L, with vertical downward as the positive direction; N is the number of pages of the three-dimensional matrix L, i is the i-th page; the elements in the two-dimensional matrix represent the depth information of different pit positions;

[0058] S2. Sort the elements in the two-dimensional matrix M according to the depth, count the number of elements corresponding to different depth values, and calculate the proportion of the number of elements corresponding to different depth values ​​relative to the total number of elements;

[0059] Specifically, in the two-dimensional matrix M, each element value represents the depth value of the corresponding position. Since the depth distribution inside the same pothole on the road surface is random and disordered, the size arrangement of each element in the two-dimensional matrix M is random. In order to analyze the depth distribution law of potholes, the method of mathematical statistics is adopted. First, it is necessary to convert unnecessary data into ordered data. Therefore, the present invention first sorts the elements in the two-dimensional matrix M according to the depth size, and records them as h1, h2,…, hi…, hn in sequence. Among them, if the size of multiple element values ​​is the same, they are only marked once: the number of elements corresponding to different depth values ​​is counted, and they are recorded as Nh1, Nh2,…, Nhn in sequence. Calculate the ratio Ti of the number of different depths to the total number S, and the calculation method is as follows:

[0060]

[0061]

[0062] Where Nhi represents the number of matrix element values ​​corresponding to the depth with sort value hi.

[0063] S3. according to the difference in users' attention to different pit depths, weights are assigned to the proportions corresponding to different depth values, and a pit depth distribution factor is calculated; the pit depth distribution factor represents the development of the pit along the depth direction;

[0064] When analyzing the pit depth information, different researchers pay different attention to the pit depths with different proportions. For example, some researchers pay more attention to the pit depths with higher frequency or proportion, while other researchers may pay more attention to the overall pit depth distribution. Therefore, the present invention introduces the concept of weight q in the proportion Ti, which is recorded as the conversion proportion Tiz,

[0065]

[0066] When q>1, it indicates that the depth values ​​with higher frequency of occurrence inside the pits are of higher concern;

[0067] When q = 1, it indicates that the depth values ​​with higher and lower frequencies inside the pits are given equal attention;

[0068] When q<1, it indicates that the attention is paid to the depth values ​​with lower frequency inside the pits.

[0069] The invention assigns weights to the depth value proportions, which helps maintenance personnel to determine the depth value of concern by considering the weight of the depth; at the same time, the use of the proportion value can eliminate the influence of the dimension, simplify the calculation, and improve the universality of the method.

[0070] The present invention introduces the pit depth distribution factor U q , represents the depth distribution of potholes. The larger the value, the deeper the overall distribution depth of the potholes. The faster the value changes, the faster the development along the depth direction. Compared with traditional pothole characterization indicators (volume, depth), this indicator takes into account the differences in road maintainers' attention to different pothole depths. Moreover, this indicator is calculated based on all collected data, and the calculation result is more accurate, which can accurately reflect the overall depth of the entire pothole.

[0071]

[0072] S4. Calculate the average area of ​​the edge matrix corresponding to the entire pit to obtain the pit expansion factor;

[0073] The pit expansion factor Z refers to the development of the pit along the plane. The larger the expansion factor, the more the pit develops along the plane. The faster the expansion factor changes, the faster the pit expands. This indicator will be used to evaluate the development of the pit in the plane.

[0074] 1) Extract the i-th page L(x, y, i) of the three-dimensional matrix L in sequence, denoted as matrix Li, i = 1, 2, 3, ..., N; extract the edge contour matrix Fi of Li, the specific method is as follows:

[0075] 2) Create matrices B and C as follows:

[0076]

[0077]

[0078] 3) Perform the following operations on matrix Li, matrix B and matrix C respectively:

[0079] Li B =Li[x,y]*B[x,y]

[0080] Li C =Li[x,y]*C[x,y]

[0081] In the formula, * represents the convolution operation, x and y are the coordinates in two directions respectively, and Li B It is the new matrix formed by the operation of the image matrix Li and the matrix B; Li C It is a new matrix formed by the operation of image matrix Li and matrix C.

[0082] 4) Calculate Li B and Li C The maximum value at the same position in the matrix is ​​used to establish the pit edge matrix Fi. The calculation process is as follows:

[0083] Fi=Max(Li B ,Li C )

[0084] 5) Calculate the area A of the pit edge matrix Fi Fi :

[0085] A Fi =∑Fi(x,y)

[0086] 6) Calculate the average area Z of the edge matrix corresponding to the entire pit, which is the expansion factor of the pit:

[0087]

[0088] As a preferred embodiment of the present invention, considering that the system may have errors due to the influence of sensor instability during the data collection process of the device, and the system error V usually obeys the Gaussian distribution W, that is:

[0089] V~W(μ,σ 2 )

[0090] μ is the mathematical expectation and σ is the standard deviation.

[0091] Therefore, the present invention adopts Gaussian distribution to correct the pit depth distribution factor and the expansion factor respectively to obtain the pit correction depth distribution factor U' q And the modified dilation factor Z':

[0092]

[0093]

[0094] S5. sequentially collect the pit 3D point cloud information corresponding to different time periods, and execute steps S1-S4 for the pit 3D point cloud information collected at each time period to obtain the pit distribution factor and pit expansion factor at the corresponding time period;

[0095] Specifically, data at different times are collected, and the corresponding pit correction depth distribution factor and correction expansion factor are calculated. The change gradient of the correction depth distribution factor and the change gradient of the correction expansion factor are calculated, and the change law of the pit size is analyzed, as follows:

[0096] 1) Collect the three-dimensional point cloud information of the pits corresponding to the time t1, t2, ..., tN in sequence, repeat steps S1-S6, and obtain the corrected pit distribution factor U' at the corresponding time q1 ,U' q2 ,…,U' qN , and the modified pit expansion factors Z'1, Z'2, ..., Z' N ; Establish the pit distribution factor change gradient dU, calculated as follows:

[0097]

[0098] Δt=t i+1 -t i

[0099] In the formula, Δt is the time interval, t i+1 and t i are respectively the i+1th moment and the ith moment, U' qi and U' q(i+1) are the modified pit distribution factors at the i-th moment and the i+1-th moment respectively.

[0100] The pit expansion factor change gradient dU is established and calculated as follows:

[0101]

[0102] S6. Based on the change gradient of the pit distribution factor and the change gradient of the pit expansion factor, the pit evolution law is obtained.

[0103] Specifically, 1) when dZ>0, dU>0, it indicates that the pits develop in both the depth direction and the plane direction;

[0104] 2) When dZ>0, dU=0, it indicates that the pits develop along the depth direction but not along the plane direction;

[0105] 3) When dZ>0 and dU<0, it indicates that the pits develop along the depth direction and shrink along the plane direction;

[0106] 4) When dZ = 0, dU > 0, it indicates that the pits and grooves are not developed along the depth direction but developed along the plane direction;

[0107] 5) When dZ = 0, dU = 0, it indicates that the pits and grooves are not developed along the depth direction and are not developed along the plane direction;

[0108] 6) When dZ = 0, dU < 0, it indicates that the pits are not developed along the depth direction and shrink along the plane direction;

[0109] 7) When dZ < 0, dU > 0, it indicates that the pits become shallower along the depth direction and develop along the plane direction;

[0110] 8) When dZ < 0, dU = 0, it indicates that the pits become shallower along the depth direction and do not develop along the plane direction;

[0111] 9) When dZ<0, dU<0, it indicates that the pit becomes shallower along the depth direction and shrinks along the plane direction.

[0112] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for analyzing the evolution law of road potholes, characterized in that: include: S1. Collecting three-dimensional point cloud information of the road surface and marking the non-pothole area and the pothole area to obtain a three-dimensional matrix L; converting the three-dimensional matrix L into a two-dimensional matrix M containing pothole depth information by plane projection; S2. Sort the elements in the two-dimensional matrix M according to the depth, count the number of elements corresponding to different depth values, and calculate the proportion of the number of elements corresponding to different depth values ​​relative to the total number of elements; S3. according to the difference in users' attention to different pit depths, weights are assigned to the proportions corresponding to different depth values, and a pit depth distribution factor is calculated; the pit depth distribution factor represents the development of the pit along the depth direction; S4. Calculate the average area of ​​the edge matrix corresponding to the entire pit to obtain the pit expansion factor; The pit expansion factor represents the development of the pit along the plane direction; S5. sequentially collect the pit 3D point cloud information corresponding to different time periods, and execute steps S1-S4 for the pit 3D point cloud information collected at each time period to obtain the pit depth distribution factor and the pit expansion factor at the corresponding time period; S6. Obtain the evolution law of the pits according to the change gradient of the pit depth distribution factor and the change gradient of the pit expansion factor; The pit depth distribution factor is: Nhi represents the number of different depth values, S represents the total number of elements in the two-dimensional matrix, q represents the weight, and h represents the i Represents the elements corresponding to different depth values; Step S4 specifically includes: extracting the i-th page L(x, y, i) of the three-dimensional matrix L in sequence to obtain the matrix Li, extracting the edge contour matrix Fi of Li; performing the following operations on the matrix Li, the matrix B and the matrix C respectively: Yes B =Li[x,y]*B[x,y] Li C =Li[x,y]*C[x,y] In the formula, * represents the convolution operation, Li B It is the new matrix formed by the operation of the image matrix Li and the matrix B; Li C It is a new matrix formed by the operation of the image matrix Li and the matrix C; the matrix B and the matrix C represent the edge operator; Calculate Li B and Li C The maximum value at the same position in the matrix is ​​used to obtain the pit edge matrix Fi; Calculate the area A of the pit edge matrix Fi Fi =ΣFi(x,y); Calculate the average area Z of the edge matrix corresponding to the entire pit and get the pit expansion factor 2. The method for analyzing the evolution law of road potholes according to claim 1, characterized in that: The two-dimensional matrix M is constructed as follows: x and y are coordinates in two directions respectively, z is the vertical axis direction of the matrix L, with vertical downward as the positive direction; N is the number of pages of the three-dimensional matrix L, i is the i-th page, i=1, 2, 3, ..., N, and N represents the number of pages of the three-dimensional matrix.

3. The method for analyzing the evolution law of road potholes according to claim 1, characterized in that: Before executing step S6, the method further includes correcting the pit depth distribution factor and the expansion factor according to the system error to obtain a pit correction depth distribution factor U' q and the pit correction expansion factor Z'.

4. The method for analyzing the evolution law of road potholes according to claim 3, characterized in that: Pothole correction depth distribution factor U' q And the pit correction expansion factor Z' is: V represents the systematic error.

5. A method for analyzing the evolution law of road potholes according to any one of claims 1 to 4, characterized in that: Step S1 includes: The three-dimensional point cloud information of the road surface is converted into a three-dimensional matrix H; and a pothole edge matrix is ​​established for each layer of the three-dimensional point cloud information of the road surface; The three-dimensional matrix H is cut by using the matrix formed by the non-zero values ​​in the pit edge matrix to form a three-dimensional matrix J representing the circumscribed cuboid of the three-dimensional pit region; The plane K parallel to the road surface is translated from top to bottom or from bottom to top to intersect with the matrix J. During the intersection process, the non-pothole area and the pothole area are marked respectively to obtain the three-dimensional matrix L.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for analyzing the evolution law of road potholes as described in any one of claims 1 to 5 is implemented.

Images