An automated detection, analysis and evaluation method for asphalt pavement defects

Through automated detection methods, the surface and deep information of the highway is obtained by taking photos and radar, high-pass filtering and comparison are carried out, and road disease maps are generated, which solves the problem of inability to detect highway diseases in the existing technology, and improves road maintenance efficiency and driving safety .

CN112348798BActive Publication Date: 2024-12-03JIANGSU MODERN ENG TESTING CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202011247347.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-11-10
Publication Date
2024-12-03
Estimated Expiration
2040-11-10

AI Technical Summary

Technical Problem

The existing technology cannot detect road diseases in a timely manner, resulting in roads being unable to be maintained in a timely manner, affecting driving safety.

Method used

Automatic detection methods are adopted to obtain surface information by taking photos along the road surface, use radar to obtain deep information, and high-pass filtering, comparison and correction of the two are generated to generate road disease maps to achieve timely detection and evaluation of highway diseases.

Benefits of technology

Timely detection and maintenance of highway diseases has been achieved, road stability has been improved, and vehicle driving safety has been ensured.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112348798B_ABST
    Figure CN112348798B_ABST
Patent Text Reader

Abstract

The present invention discloses an automated detection and analysis and evaluation method for asphalt pavement defects, including taking photos along the road surface to obtain surface information; using radar to obtain deep information of the road surface; uploading the surface information and deep information to a control center; comparing the surface information with a surface database to obtain the defect type; comparing the deep information with a deep database to obtain the hazard level; and generating a pavement defect map using the defect type and hazard level together. The defect type is obtained through surface information, the hazard level is obtained through deep information, and the pavement defect types and hazard levels of all time periods are combined to generate a total pavement defect map, so that the pavement damage of the detected section can be grasped as a whole, and areas with higher hazard levels can be repaired in a timely manner, thereby improving the stability of the road surface and making vehicle driving safer, thereby solving the problem that the existing method cannot detect highway defects in a timely manner, so that the highway cannot be maintained in a timely manner and affects driving.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of highway disease detection, and in particular to an automated detection and analysis and evaluation method for asphalt pavement diseases. Background Art

[0002] In recent years, my country's transportation infrastructure construction, as a strategic goal of national priority development, has achieved rapid growth in total scale. With the continuous growth of highway mileage and the rapid increase in traffic volume, public demand has also increased. Therefore, how to do a good job in maintenance and keep the highway in a good operating state is a very arduous task facing highway maintenance managers.

[0003] The coring method in the prior art mainly involves taking samples on site to test the indoor volume and physical indicators, which makes it impossible to detect the situation on the highway site, making it impossible to grasp the highway conditions in time, causing highway damage and affecting driving. Summary of the invention

[0004] The purpose of the present invention is to provide an automated detection and analysis and evaluation method for asphalt pavement defects, aiming to solve the problem that existing methods are unable to detect highway defects in a timely manner, resulting in the inability to maintain the highway in a timely manner and affecting driving.

[0005] To achieve the above objectives, on the one hand, the present invention provides an automated detection, analysis and evaluation method for asphalt pavement defects, comprising taking photos along the road surface to obtain surface information; using radar to obtain deep road surface information; uploading the surface information and deep information to a control center; comparing the surface information with a surface database to obtain the type of defect; comparing the deep information with a deep database to obtain the hazard level; and generating a road surface defect map using the defect type and hazard level together.

[0006] Before uploading the surface information and the deep information to the control center, the method further includes: performing high-pass filtering on the deep information.

[0007] Among them, the specific steps of high-pass filtering the deep information are: performing low-pass filtering and full-spectrum fidelity processing; and differentiating the full-spectrum section numerical image obtained by the full-spectrum fidelity processing and the low-pass filtered continuous section numerical image obtained by the low-pass filtering processing.

[0008] Before generating a road surface disease map by combining the disease type and the hazard level, the method further includes: recording vibration information along the road surface and correcting the hazard level.

[0009] Among them, the specific steps of taking photos along the road surface to obtain surface information are: taking photos of the road surface to obtain a grayscale image; cutting the grayscale image and extracting a black and white binary image containing edge information, where the edge is white and the background is black; filling the lines and contour gaps of the black and white binary image to obtain surface information.

[0010] The specific steps of using radar to obtain deep road surface information are: using array radar to scan along the road surface; receiving road surface reflection information; storing and splicing the reflection information to form deep information.

[0011] The scanning interval is 1-2S.

[0012] The present invention provides an automated detection, analysis and evaluation method for asphalt pavement defects, including taking photos along the road surface to obtain surface information; using radar to obtain deep road surface information; uploading the surface information and deep information to a control center; comparing the surface information with a surface database to obtain the defect type; comparing the deep information with a deep database to obtain the hazard level; and generating a pavement defect map using the defect type and hazard level together. The defect type can be obtained by comparing the surface information with the surface database, and then the hazard level can be obtained by comparing the deep information with the deep database. Finally, the pavement defect types and hazard levels of all time periods are used to generate a total pavement defect map, so that the pavement damage of the detected section can be displayed and grasped as a whole, so that areas with higher hazard levels can be repaired in a timely manner, the stability of the road surface can be improved, and vehicle driving can be made safer, thereby solving the problem that the existing method cannot detect highway defects in a timely manner, so that the highway cannot be maintained in a timely manner and driving is affected. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0014] Figure 1 It is a flow chart of an automated detection and analysis and evaluation method for asphalt pavement defects of the present invention;

[0015] Figure 2 It is a flow chart of taking photos along the road surface to obtain surface information of the present invention;

[0016] Figure 3 is a flow chart of using radar to obtain road surface deep layer information of the present invention;

[0017] Figure 4 is a flow chart of high-pass filtering of deep information according to the present invention;

[0018] Figure 5 It is a flow chart of the present invention for comparing surface information with a surface database to obtain the disease type;

[0019] Figure 6 It is a flow chart of the present invention for comparing deep information with a deep database to obtain a hazard level. DETAILED DESCRIPTION

[0020] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0021] See also Figure 1 The present invention provides an automated detection and analysis and evaluation method for asphalt pavement defects, comprising:

[0022] S101 takes photos along the road surface to obtain surface information;

[0023] Place a test vehicle on the road to be tested, place a camera array on the test vehicle, and arrange it along the cross section of the road so that a cross section of the road can be photographed at the same time to reduce distortion caused by camera distortion. Figure 2 , the specific steps are:

[0024] S201 takes a photo of the road surface to obtain a grayscale image;

[0025] Grayscale images are images that divide white and black into several levels according to a logarithmic relationship. Grayscale is divided into 256 levels. The road damage in the image can be identified by the grayscale value change rate of the image pixels.

[0026] S202 cuts the grayscale image and extracts a black-and-white binary image containing edge information, where the edge is white and the background is black;

[0027] The grayscale image is cut into blocks according to a certain length, and the areas with large grayscale value change rates are marked and edge information is identified. The edges are set to white and the others are set to black, so that the white areas are feature areas.

[0028] S203 fills the gaps between lines and contours of the black and white binary image to obtain surface information.

[0029] The breakpoints are filled based on the interpolation method to form a continuous white outline image. The optional interpolation functions include polynomials and trigonometric functions. For values ​​with large value differences, segmented interpolation can also be performed to obtain the surface information of the current area.

[0030] The S102 uses radar to obtain deep information about the road surface;

[0031] The camera method will introduce large errors due to the influence of road debris and the camera, so radar scanning is used for correction. The radar emits electromagnetic waves to illuminate the target and receives its echo, thereby obtaining information such as the distance from the target to the electromagnetic wave emission point, the distance change rate (radial velocity), direction, and altitude. When irradiating debris on the surface, the distance is shorter, and when irradiating gaps and end faces, the reflection distance is longer, so that the depth of the gap can be judged and the interference of objects above the road surface can be eliminated.

[0032] See also Figure 3 , the specific steps are:

[0033] The S301 uses array radar to scan along the road surface;

[0034] The array radar is placed along the cross section of the road surface, and by scanning a cross section of the road surface at the same time, the road surface depth at the corresponding position can be obtained, and the data of the adjacent radar can also be corrected. Considering the power consumption requirements, the scanning interval is 1-2S.

[0035] S302 receives road surface reflection information;

[0036] Use the radar receiving system to receive the reflected information from the road surface.

[0037] S303 stores and splices the reflection information to form deep information.

[0038] The reflection information of all radars is stored, then all the reflection information is stitched together based on a pre-designed stitching algorithm, and the parameters of adjacent radars are corrected to form the final deep information.

[0039] S103 performs high-pass filtering on the deep information;

[0040] Radar will produce errors in the process of receiving data, so filtering is required.

[0041] See also Figure 4 , the specific steps are:

[0042] S401 performs low-pass filtering and full-spectrum fidelity processing;

[0043] S402 performs a differential operation on the full spectrum cross-sectional numerical image obtained by the full spectrum fidelity processing and the low-pass filtered continuous cross-sectional numerical image obtained by the low-pass filtering processing.

[0044] S104 uploads the surface information and deep information to the control center;

[0045] Processing data requires powerful computing power, but it is expensive to install such a system on the measurement vehicle, so cloud computing is used to upload the information to the control center for processing.

[0046] S105 compares the surface information with the surface database to obtain the disease type;

[0047] See also Figure 5 , the specific steps are:

[0048] S501 obtains a surface database;

[0049] Before the detection, a database of some surface information features has been collected in advance, and different labels have been extracted for different disease characteristics, so that they can be compared with the current detection data to confirm the existing disease type.

[0050] S502 performs expansion detection on the surface information based on the surface database and marks the determined diseased area;

[0051] A large amount of test data will be generated in the entire section of the road to be tested. In order to make a more accurate judgment, an expansion test is used to comprehensively judge the surface information at the current moment and the surface information at nearby moments, so that the damage condition of a section of the road can be judged.

[0052] S503 connects the diseased areas to form an overall disease map;

[0053] By connecting adjacent diseased areas, an overall disease map can be formed to assist in judgment.

[0054] S504 determines the disease type of the overall disease map based on the surface database.

[0055] The labels of the overall damage map are extracted and compared with the surface database to determine the type of damage on the current road section.

[0056] S106 compares the deep information with the deep database to obtain the hazard level;

[0057] See also Figure 6 , the specific steps are:

[0058] S601 segments the deep information;

[0059] The deep information of the same cross section is segmented along the cross section, so that the disease conditions of different sections can be compared simultaneously to improve work efficiency.

[0060] S602 extracts the maximum grayscale value of each segment of deep information;

[0061] The maximum grayscale value of each segment of deep information represents the deepest depth, so that the hazard level of the current segment can be judged. The deeper the gap, the larger the maximum grayscale value.

[0062] S603 compares the maximum grayscale value with the deep database to obtain the hazard level of each segment;

[0063] Different maximum grayscale values ​​are pre-segmented in the deep database, and the hazard level can be obtained by comparing the maximum grayscale value of each segment of deep information with the deep database.

[0064] S604 calculates the total hazard level based on different weights.

[0065] Since the rolling conditions of the road are inconsistent, vehicles usually drive close to the middle of the road, so there are two positions that are often compressed. Therefore, the road can be divided into five sections. The two ends of the middle section are under heavier pressure. The calculation formula that can be used is as follows:

[0066] A=0.1A1+0.3A2+0.2A3+0.3A4+0.1A5, where A is the total hazard level, A i is the hazard level of each segment.

[0067] S107 records vibration information along the road surface and corrects the hazard level;

[0068] A gyroscope installed on the detection vehicle can record the vehicle's vibration. Where the vibration amplitude is large, the road hazard level is higher, so the hazard level generated by the radar can be further corrected to improve accuracy.

[0069] S108 generates a pavement damage map by combining the damage type and the hazard level.

[0070] Finally, by generating a total pavement damage map based on the pavement disease types and hazard levels of all time periods, the pavement damage conditions of the inspected sections can be displayed and grasped as a whole, so that timely repairs can be carried out in areas with higher hazard levels, improving the stability of the road surface and making vehicle driving safer, thus solving the problem that existing methods are unable to detect highway diseases in a timely manner, resulting in the inability to maintain the highway in a timely manner and affecting driving.

[0071] What is disclosed above is only a preferred embodiment of the present invention, and it certainly cannot be used to limit the scope of rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiment and equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. An automated detection and analysis and evaluation method for asphalt pavement defects, characterized in that: This includes taking photos along the road surface to obtain surface information; Use radar to obtain deep information about the road surface; Upload surface information and deep information to the control center; Compare the surface information with the surface database to obtain the disease type; Compare the deep information with the deep database to obtain the hazard level; Generate a pavement damage map by combining the damage type and damage level; Before uploading the surface layer information and the deep layer information to the control center, the method further includes: performing high pass filtering on the deep layer information; The specific steps of high-pass filtering the deep information are: Perform low-pass filtering and full spectrum fidelity processing; The full spectrum cross-section numerical image obtained by full spectrum fidelity processing is differentiated from the low-pass filtered continuous cross-section numerical image obtained by low-pass filtering processing; Compare the surface information with the surface database to obtain the disease type. The specific steps are as follows: Before the test, a database of some surface information features has been collected in advance, and different labels are extracted for different disease features, which are compared with the current test data to confirm the existing disease type; The surface information at the current moment and the surface information at the nearby moments are comprehensively judged by expansion detection, and the judged diseased area is marked; Connect adjacent diseased areas to form an overall disease map to assist in judgment; Extract the labels of the overall damage map and compare them with the surface database to determine the type of damage on the current road section; Compare the deep information with the deep database to obtain the hazard level. The specific steps are as follows: For the deep information of the same cross section, the cross section is divided into sections, and the disease conditions of different sections are compared simultaneously; Extract the maximum grayscale value of each segment of deep information; Compare the maximum grayscale value with the deep database to obtain the hazard level of each segment; The total hazard level is calculated based on different weights. Since the road surface rolling conditions are inconsistent, vehicles generally drive in the middle of the road surface, which results in two positions being frequently compressed. The road surface is divided into five sections, and the two adjacent ends of the middle section are more compressed. The calculation formula used is as follows: A=0.1A1+0.3A2+0.2A3+0.3A4+0.1A5, where A is the total hazard level, A i is the hazard level of each segment; The vibration information is recorded along the road surface and the hazard level is corrected.

2. The method for automatic detection and analysis and evaluation of asphalt pavement defects according to claim 1, characterized in that: The specific steps of taking photos along the road surface to obtain surface information are: Take a photo of the road surface to obtain a grayscale image; Cut the grayscale image and extract the black-and-white binary image containing edge information, where the edge is white and the background is black; Fill the gaps between black and white binary image lines and contours to obtain surface information.

3. The method for automatic detection, analysis and evaluation of asphalt pavement defects according to claim 1, characterized in that: The specific steps of using radar to obtain deep road surface information are: Use array radar to scan along the road surface; Receive road surface reflection information; The reflection information is stored and spliced ​​to form deep information.

4. The method for automatic detection, analysis and evaluation of asphalt pavement defects according to claim 3, characterized in that: The scanning interval is 1-2S.