Disclosure of Invention
In order to solve the problems, the invention provides a highway construction quality automatic detection method and system.
The invention discloses a highway construction quality automatic detection method and a system, which adopt the following technical scheme:
the embodiment of the invention provides a highway construction quality automatic detection method, which comprises the following steps:
acquiring a road edge image of road construction completion;
Acquiring an edge area corresponding to each linear edge in the road edge image; dividing the edge areas to obtain a plurality of block areas of each edge area; acquiring a priori gray value of the lane line according to the gray difference of the adjacent block areas in the edge area; obtaining the abnormal degree of each linear edge in the road edge image according to the difference between the gray value corresponding to the pixel point on the linear edge and the priori gray value of the lane line;
The continuous edges in the road edge images are all used as suspected crack edges, and the direction consistency degree of two adjacent straight line edges in each suspected crack edge is obtained according to the included angle value of the two adjacent straight line edges in the suspected crack edges; obtaining the regularity of each suspected crack edge in the road edge image according to the direction consistency degree and the flatness of the suspected crack edge;
obtaining the final abnormality degree of each linear edge in the road edge image according to the abnormality degree of the linear edge and the regularity of the suspected crack edge where the linear edge is located; and screening crack edges in the road edge image according to the final degree of abnormality, and determining a crack region of the road surface through the crack edges.
Further, the step of obtaining the lane line prior gray value according to the gray difference of the adjacent block areas in the edge area comprises the following specific steps:
For an edge region corresponding to any one straight line edge in a road edge image, acquiring absolute differences of gray average values corresponding to two adjacent block regions in the edge region, acquiring absolute differences of gray value range corresponding to the two adjacent block regions, and fusing the absolute differences of the gray average values and the absolute differences of the gray value range to obtain gray difference indexes of the two adjacent block regions;
Obtaining a comprehensive gray level change index of each edge region according to gray level difference indexes of adjacent partitioned regions in the edge region;
and acquiring the prior gray value of the lane line according to the magnitude of the comprehensive gray change index.
Further, the method for obtaining the comprehensive gray scale change index of each edge region according to the gray scale difference index of the adjacent partitioned regions in the edge region comprises the following specific steps:
And for an edge region corresponding to any one straight line edge in the road edge image, acquiring the accumulated value of the gray difference indexes of each block region and all adjacent block regions around in the edge region, and taking the sum of the accumulated values of the gray difference indexes corresponding to all block regions in the edge region as the comprehensive gray change index of the edge region.
Further, the step of obtaining the priori gray value of the lane line according to the magnitude of the comprehensive gray change index comprises the following specific steps:
And acquiring a comprehensive gray scale change index of an edge area corresponding to each linear edge in the road edge image, and taking a gray scale average value of the edge area corresponding to the minimum value of the comprehensive gray scale change index as a lane line prior gray scale value.
Further, the obtaining the abnormal degree of each straight line edge in the road edge image according to the difference between the gray value corresponding to the pixel point on the straight line edge and the prior gray value of the lane line comprises the following specific steps:
And for any straight line edge in the road edge image, acquiring the absolute difference value of the gray value corresponding to each pixel point on the straight line edge and the prior gray value of the lane line, taking the average value of the absolute differences of the gray values corresponding to all the pixel points on the straight line edge and the prior gray value of the lane line as the initial abnormal degree of the straight line edge, and carrying out normalization processing on the initial abnormal degree to obtain the abnormal degree of the straight line edge.
Further, the obtaining the direction consistency degree of the two adjacent straight edges in each suspected crack edge according to the included angle value of the two adjacent straight edges in the suspected crack edge comprises the following specific steps:
Marking any suspected crack edge in the road edge image as a target suspected crack edge; and taking an inverse proportion value of the minimum included angle value of two adjacent straight line edges in the target suspected crack edges as the direction consistency degree of the two adjacent straight line edges.
Further, the step of obtaining the regularity of each suspected crack edge in the road edge image according to the direction consistency degree and the flatness of the suspected crack edge comprises the following specific steps:
Taking the ratio of the number of the pixels of all the straight edges in the target suspected crack edge to the number of the pixels in the target suspected crack edge as the flatness of the target suspected crack edge; and fusing and normalizing the accumulated values of the flatness and the direction consistency degree of all adjacent straight line edges in the target suspected crack edges to obtain the regularity of the target suspected crack edges.
Further, the obtaining the final abnormality degree of each linear edge in the road edge image according to the abnormality degree of the linear edge and the regularity of the suspected crack edge where the linear edge is located comprises the following specific steps:
For any one straight line edge in the road edge image; fusing and normalizing the abnormality degree of the linear edge and the regularity of the suspected crack edge where the linear edge is positioned to obtain the final abnormality degree of the linear edge; the regularity is inversely related to the final degree of abnormality.
Further, the specific acquiring method of the edge area is as follows:
for any one straight line edge in the road edge image, acquiring one straight line edge closest to the straight line edge, and taking a quadrangle formed by the straight line edge and the straight line edge closest to the straight line edge as an edge area corresponding to the straight line edge.
The invention also provides an automatic road construction quality detection system, which comprises a memory and a processor, wherein the processor executes a computer program stored in the memory so as to realize the steps of the method.
The technical scheme of the invention has the beneficial effects that: after the road edge image with the road construction completed is obtained, the lane line prior gray value is obtained by analyzing the gray difference of the adjacent block areas in the edge area, and the abnormality degree of each linear edge in the road edge image is obtained by analyzing the difference of the gray value corresponding to the pixel point on the linear edge and the lane line prior gray value, so that the accuracy of identifying the linear edge formed by the road surface crack is improved; and combining the direction consistency degree of two adjacent linear edges in the whole suspected crack edge where the linear edge is positioned and the straightness condition of the suspected crack edge to obtain the regularity of the suspected crack edge in the road edge image, accurately identifying the crack edge in the road edge image by combining the abnormality degree and the regularity, further determining the crack area, solving the problem of inaccurate crack detection result caused by the similarity of partial cracks on the road surface and the lane line edges on the road surface, and improving the rationality of automatic detection of the road construction quality.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following description refers to the specific implementation, structure, characteristics and effects of the automatic detection method and system for highway construction quality according to the present invention by combining the accompanying drawings and the preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The invention provides a method and a system for automatically detecting the construction quality of a highway.
Referring to fig. 1, a flowchart of steps of a method for automatically detecting highway construction quality according to an embodiment of the present invention is shown, the method includes the following steps:
And S001, acquiring a road edge image of the road construction completion.
It should be noted that, the main purpose of this embodiment is to determine the edge formed by the crack in the road edge image after the road construction is completed, so as to better determine the crack area, complete the automatic detection of the road construction quality, and first obtain the relevant data before starting the analysis.
Specifically, a road edge image of the road construction completion is obtained, specifically as follows:
Carrying out overlooking shooting on the road surface by an unmanned aerial vehicle carrying a high-resolution CCD camera to obtain an RGB image of the road surface; carrying out graying treatment on the RGB image of the road surface to obtain a road gray image; carrying out Canny edge detection on the road gray level image to obtain a road edge initial image; and performing open operation on the initial image of the road edge to obtain the image of the road edge.
The method is characterized in that the initial image of the road edge contains some small noise or isolated points, so that the local noise is removed by performing open operation on the initial image of the road edge to obtain the image of the road edge; referring to fig. 2, fig. 2 is a road gray scale image of the present embodiment; referring to fig. 3, fig. 3 is a road edge image of the present embodiment; the operation is performed on the initial image of the road edge, and the obtained image of the road edge is a morphological existing method, which is not described in detail in this embodiment.
Thus, road edge images of road construction completion are obtained.
Step S002, obtaining an edge area corresponding to each linear edge in the road edge image; dividing the edge areas to obtain a plurality of block areas of each edge area; acquiring a priori gray value of the lane line according to the gray difference of the adjacent block areas in the edge area; and obtaining the abnormal degree of each linear edge in the road edge image according to the difference between the gray value corresponding to the pixel point on the linear edge and the priori gray value of the lane line.
It should be noted that, the main purpose of this embodiment is to determine the edge formed by the crack in the road edge image after the road construction is completed, so as to better determine the crack area, complete the automatic detection of the road construction quality, and to facilitate the subsequent better analysis, it is preferred to obtain the straight line edge in the road edge image and the edge area formed by the straight line edge.
Specifically, a plurality of straight edges in the road edge image are obtained, and the method specifically comprises the following steps:
And carrying out Hough straight line detection on edges in the road edge image to obtain a plurality of straight line edges in the road edge image.
The present embodiment is not described in detail, and the present embodiment is used for performing hough line detection on edges in a road edge image to obtain a conventional method for detecting that a plurality of straight edges in the road edge image are hough lines.
Further, an edge area corresponding to each straight line edge in the road edge image is obtained, which is specifically as follows:
for any one straight line edge in the road edge image, acquiring one straight line edge closest to the straight line edge, and taking a quadrangle formed by the straight line edge and the straight line edge closest to the straight line edge as an edge area corresponding to the straight line edge.
It should be noted that, when the straight line edge is a line segment, in this embodiment, when obtaining a straight line edge with the closest straight line edge, the midpoint of the straight line edge is selected, then the perpendicular line distance between the midpoint and the adjacent straight line edge is obtained, and the adjacent straight line edge with the shortest perpendicular line distance is used as the straight line edge with the closest straight line edge distance.
It should be noted that, in order to facilitate the subsequent better analysis of the gray level change of the edge region, the edge region needs to be divided into small regions for local analysis.
Specifically, the edge areas are divided to obtain a plurality of block areas of each edge area, which is specifically as follows:
For an edge area corresponding to any one straight edge in the road edge image, dividing the edge area into a plurality of partsIs provided with a partition area of the (c),The block proportion coefficient is a preset block proportion coefficient, and the block proportion coefficient is specifically 3.
It should be noted that the normal block area size of the edge area isIf the division into the last remaining area is notThe present embodiment does not act as a partitioned area.
It should be noted that, because the color of the lane line is artificially colored, generally white or yellow, the gray level distribution of the area is relatively uniform, the gray level overall change is not large, the crack is naturally formed, the gray level distribution of the area is relatively discrete, and the gray level overall change is large, therefore, the gray level difference index of the adjacent partitioned areas is obtained by analyzing the gray level change of the adjacent partitioned areas in the edge area, and the larger the gray level difference index, the more likely the current edge area is the crack area.
Specifically, according to the gray level difference of the adjacent block areas in the edge area, the prior gray level value of the lane line is obtained, and the method comprises the following steps:
firstly, according to the gray level difference of the adjacent partitioned areas in the edge areas, gray level difference indexes of the adjacent partitioned areas in each edge area are obtained.
And secondly, obtaining the comprehensive gray scale change index of each edge region according to the gray scale difference index of the adjacent partitioned regions in the edge region.
Finally, according to the magnitude of the comprehensive gray level change index, the prior gray level value of the lane line is obtained.
Preferably, in one embodiment of the present invention, according to the gray scale difference of the adjacent segmented regions in the edge region, the gray scale difference index of the adjacent segmented region in each edge region is obtained, which is specifically as follows:
And for an edge region corresponding to any one straight line edge in the road edge image, acquiring absolute differences of gray average values corresponding to two adjacent block regions in the edge region, acquiring absolute differences of gray value range corresponding to the two adjacent block regions, and fusing the absolute differences of the gray average values and the absolute differences of the gray value range to obtain gray difference indexes of the two adjacent block regions.
As a specific example, the specific method for obtaining the gray scale difference index is as follows:
And marking an edge area corresponding to any one straight edge in the road edge image as a target edge area.
In the formula,Is the first of the target edge areaGray average values corresponding to the partitioned areas; Is the first of the target edge areaAdjacent ones of the partitioned areasGray average values corresponding to the partitioned areas; Is the first of the target edge areaThe pixel points in the individual block areas correspond to the extremely poor gray values; Is the first of the target edge areaAdjacent ones of the partitioned areasThe pixel points in the individual block areas correspond to the extremely poor gray values; Taking an absolute value; Is the first in the edge area of the objectThe block areas are adjacent to the firstGray scale difference index of each block region.
It should be noted that, because the color of the lane line is artificially colored, generally white or yellow, the gray level distribution of the area is relatively uniform, the gray level overall change is not large, the crack is naturally formed, the gray level distribution of the area is relatively discrete, and the gray level overall change is large, therefore, the formula obtains the gray level difference index of the adjacent segmented area by analyzing the gray level change of the adjacent segmented area in the edge area,The absolute difference of the corresponding gray average values of the two adjacent partitioned areas is reflected,The absolute difference of the corresponding gray value range of the two adjacent partitioned areas is reflected, and the larger the gray difference index is, the more likely the current edge area is a crack area.
The gray level difference of two adjacent block areas in one edge area is obtained, and the gray level difference of all adjacent block areas in the same edge area is analyzed to obtain the comprehensive gray level change index of the edge area, so that the larger the comprehensive gray level change index of the edge area is, the more likely the current edge area is a crack area.
Preferably, in one embodiment of the present invention, according to the gray level difference index of the adjacent partitioned areas in the edge area, the comprehensive gray level change index of each edge area is obtained, which is specifically as follows:
And for an edge region corresponding to any one straight line edge in the road edge image, acquiring the accumulated value of the gray difference indexes of each block region and all adjacent block regions around in the edge region, and taking the sum of the accumulated values of the gray difference indexes corresponding to all block regions in the edge region as the comprehensive gray change index of the edge region.
As a specific example, the specific method for obtaining the comprehensive gray scale variation index is as follows:
In the formula,The number of the blocking areas in the target edge area; Is the first in the target edge areaAccumulated values of gray difference indexes of each block area and all adjacent block areas around; Is an integrated gray scale change index of the target edge area.
The method is characterized in that the gray level distribution of the lane line area is uniform, the gray level overall change is not large, the gray level distribution of the area formed by the crack is discrete, the gray level overall change is large, the gray level difference of the adjacent partitioned areas in one edge area is analyzed, the comprehensive gray level change index of the whole edge area is obtained by analyzing the gray level difference of all the adjacent partitioned areas in the same edge area, and the more the comprehensive gray level change index is large, the more likely the current edge area is the crack area is.
The above-mentioned comprehensive gray scale change index of the edge region is obtained, and in an ideal case, the comprehensive gray scale change index of the lane line region should be very small and basically unchanged, so that the critical lane line prior gray scale value is obtained by selecting the edge region with the minimum comprehensive gray scale change index as the lane line reference region.
Specifically, according to the magnitude of the comprehensive gray level change index, the prior gray level value of the lane line is obtained, specifically as follows:
And acquiring a comprehensive gray scale change index of an edge area corresponding to each linear edge in the road edge image, and taking a gray scale average value of the edge area corresponding to the minimum value of the comprehensive gray scale change index as a lane line prior gray scale value.
It should be noted that, the lane line prior gray value is obtained, the degree of abnormality of the straight line edge in the road edge image is analyzed by combining the lane line prior gray value, when the gray value corresponding to the pixel point on the straight line edge is not greatly different from the lane line prior gray value, the more likely that the straight line edge is the edge of the lane line is, the more likely that the straight line edge is the straight line edge formed by the crack is, and the greater the degree of abnormality is.
Preferably, in one embodiment of the present invention, the degree of abnormality of each straight line edge in the road edge image is obtained according to the difference between the gray value corresponding to the pixel point on the straight line edge and the prior gray value of the lane line, which is specifically as follows:
And for any straight line edge in the road edge image, acquiring the absolute difference value of the gray value corresponding to each pixel point on the straight line edge and the prior gray value of the lane line, taking the average value of the absolute differences of the gray values corresponding to all the pixel points on the straight line edge and the prior gray value of the lane line as the initial abnormal degree of the straight line edge, and carrying out normalization processing on the initial abnormal degree to obtain the abnormal degree of the straight line edge.
As a specific example, the specific method for obtaining the abnormality degree is as follows:
In the formula,Is the first in the road edge imageThe number of pixels on each straight edge; Is the first in the road edge imageOn the straight edgeGray values corresponding to the pixel points; The prior gray value of the lane line is obtained; Taking an absolute value; Is a linear normalization function used for normalization; Is the first in the road edge imageDegree of abnormality of the individual straight edges.
The method is characterized in that when the gray value corresponding to the pixel point on the linear edge is not greatly different from the priori gray value of the lane line, the more likely the linear edge is the edge of the lane line, the smaller the degree of abnormality is; whereas the more likely the straight edge is a straight edge for crack formation, the greater the degree of anomaly; the above formula obtains the degree of abnormality of the linear edge by analyzing the difference between the gray values corresponding to all the pixel points on the linear edge and the priori gray values of the lane lines, and the larger the degree of abnormality is, the more likely the linear edge is a linear edge formed by cracks.
Thus, the degree of abnormality of each straight line edge in the road edge image is obtained.
Step S003, using continuous edges in the road edge image as suspected crack edges, and obtaining the direction consistency degree of two adjacent straight edges in each suspected crack edge according to the included angle value of the two adjacent straight edges in the suspected crack edges; and obtaining the regularity of each suspected crack edge in the road edge image according to the direction consistency degree and the flatness of the suspected crack edge.
It should be noted that, the degree of abnormality of the straight line edge is obtained only by analyzing the change of the gray value corresponding to the pixel point on the single straight line edge, and because the gray value of the straight line edge formed by partial cracks is not greatly changed compared with the prior gray value of the lane line, it is difficult to accurately determine whether the straight line edge is abnormal, and therefore, the integral analysis is also required to be performed by combining the edge where the straight line edge is located, so that the crack detection result is more accurate.
It should be further noted that, there may be multiple straight edges on the same edge in the road edge image, if the straight edges are oriented consistently, the more likely the edge where the straight edge is located is a lane line edge, so by analyzing the angle change of the adjacent straight edges on the edge where the straight edge is located, the more consistent the direction of the adjacent straight edges in the edge is obtained, and the more likely the current edge is a lane line edge.
Preferably, in one embodiment of the present invention, continuous edges in the road edge image are all used as suspected crack edges, and the direction consistency degree of two adjacent straight edges in each suspected crack edge is obtained according to the included angle value of two adjacent straight edges in the suspected crack edges, specifically as follows:
Marking any suspected crack edge in the road edge image as a target suspected crack edge; and taking an inverse proportion value of the minimum included angle value of two adjacent straight line edges in the target suspected crack edges as the direction consistency degree of the two adjacent straight line edges.
As a specific example, the specific method for obtaining the direction consistency degree is as follows:
In the formula,Is the first of the suspected crack edges of the targetThe included angle value between each straight edge and the horizontal line; The acquisition method of (1) is as follows: along one side of the edge of the target suspected crack, obtaining the firstAdjacent ones of the straight edges will beThe angle value between the adjacent straight edge of each straight edge and the horizontal line is recorded as；Taking an absolute value; the present embodiment uses an exponential function based on natural constantsIs presented in inverse proportion to the model of (c),Is the input of the model; Is the first of the suspected crack edges of the targetThe degree to which each linear edge coincides with the direction of the corresponding adjacent linear edge.
It should be noted that, there may be multiple straight line edges on the same suspected crack edge, if the directions of the adjacent straight line edges are consistent, the suspected crack edge where the straight line edge is located is more likely to be a lane line edge, so the above formula obtains the consistent degree of the directions of the two adjacent straight line edges by analyzing the included angle value of the two adjacent straight line edges, and the more consistent degree of the directions is, the more likely the current suspected crack edge is a lane line edge.
It should be noted that, the above-mentioned degree of direction coincidence of two adjacent straight edges in the suspected crack edges is obtained, if the degree of direction coincidence of all straight edges in the suspected crack edges is higher, and meanwhile, the ratio of the straight edges is also higher, the regularity of the suspected crack edges is higher, a straight line is presented, and the more likely the current suspected crack edge is a lane line edge.
Preferably, in one embodiment of the present invention, the regularity of each suspected crack edge in the road edge image is obtained according to the direction consistency degree and the flatness of the suspected crack edge, which is specifically as follows:
Taking the ratio of the number of the pixels of all the straight edges in the target suspected crack edge to the number of the pixels in the target suspected crack edge as the flatness of the target suspected crack edge; and fusing and normalizing the accumulated values of the flatness and the direction consistency degree of all adjacent straight line edges in the target suspected crack edges to obtain the regularity of the target suspected crack edges.
As a specific example, the specific method for obtaining the regularity is as follows:
In the formula,The number of straight edges in the suspected crack edges of the target; Is the first of the suspected crack edges of the targetThe degree of direction coincidence of each linear edge and the corresponding adjacent linear edge; The number of all linear edge pixel points in the target suspected crack edge is the number of all linear edge pixel points; The number of pixel points in the edge of the target suspected crack; is a softmax function for normalization; Is the regularity of the edge of the suspected crack of the target.
It should be noted that if the degree of direction coincidence of all the straight line edges in the suspected crack edges is high, and meanwhile, the ratio of the straight line edges is also high, the regularity of the suspected crack edges is high, one straight line is presented, the more likely the current suspected crack edge is a lane line edge, therefore, the above formula obtains the regularity of the target suspected crack edges by analyzing the degree of direction coincidence of all the adjacent straight line edges in the target suspected crack edges and the flatness of the suspected crack edges,The straightness of the target suspected crack edge is reflected, and the higher the regularity is, the more likely the target suspected crack edge is the edge of the lane line.
So far, the regularity of each suspected crack edge in the road edge image is obtained.
Step S004, obtaining the final abnormality degree of each linear edge in the road edge image according to the abnormality degree of the linear edge and the regularity of the suspected crack edge where the linear edge is located; and screening crack edges in the road edge image according to the final degree of abnormality, and determining a crack region of the road surface through the crack edges.
The degree of abnormality of each straight line edge in the road edge image and the degree of regularity of each suspected crack edge in the road edge image are analyzed respectively, and the final degree of abnormality of each straight line edge in the road edge image is obtained comprehensively by combining the degree of abnormality of the straight line edge and the degree of regularity of the suspected crack edge where the straight line edge is located.
Preferably, in one embodiment of the present invention, the final abnormality degree of each linear edge in the road edge image is obtained according to the abnormality degree of the linear edge and the regularity of the suspected crack edge where the linear edge is located, which is specifically as follows:
For any one straight line edge in the road edge image; fusing and normalizing the abnormality degree of the linear edge and the regularity of the suspected crack edge where the linear edge is positioned to obtain the final abnormality degree of the linear edge; the regularity is inversely related to the final degree of abnormality.
As a specific example, the specific method for obtaining the final abnormality degree is as follows:
In the formula,Is the first in the road edge imageDegree of abnormality of the individual straight edges; Is the first in the road edge imageThe regularity of the suspected crack edges where the straight edges are located; Is a preset super parameter, and aims to prevent the denominator from being 0; Is a linear normalization function used for normalization; Is the first in the road edge imageFinal degree of anomaly of the individual straight edges.
The greater the degree of abnormality of the straight line edge in the road edge image, the more likely the straight line edge is crack formation, and the greater the final degree of abnormality; meanwhile, the analysis is carried out by combining the regularity of the suspected crack edges where the straight line edges are located, if the regularity of the suspected crack edges is smaller, the abnormal straight line edges in the suspected crack edges are more likely to form cracks.
The final abnormality degree of each straight line edge in the road edge image is obtained, and the determination is performed by setting an appropriate threshold value to determine the crack edge of the road surface, thereby obtaining the crack region.
Specifically, the crack edges in the road edge image are screened according to the final degree of abnormality, and the crack areas of the road surface are determined through the crack edges, specifically as follows:
Presetting a first threshold value, describing the embodiment by taking the first threshold value as 0.7, and taking the linear edge with the final abnormality degree larger than the first threshold value as a crack linear edge, otherwise, not taking the linear edge as the crack linear edge; and taking the suspected crack edge where the crack straight line edge is positioned as a crack edge, setting the gray values of the edges except the crack edge in the road edge image as 0, obtaining a crack edge image, and taking the area formed by the crack edge in the crack edge image as a crack area of the road surface.
Through the steps, the automatic detection method for the construction quality of the highway is completed.
Another embodiment of the present invention provides an automatic road construction quality detection system, the system including a memory and a processor, the processor executing a computer program stored in the memory, performing the following operations:
Acquiring a road edge image of road construction completion; acquiring an edge area corresponding to each linear edge in the road edge image; dividing the edge areas to obtain a plurality of block areas of each edge area; acquiring a priori gray value of the lane line according to the gray difference of the adjacent block areas in the edge area; obtaining the abnormal degree of each linear edge in the road edge image according to the difference between the gray value corresponding to the pixel point on the linear edge and the priori gray value of the lane line; the continuous edges in the road edge images are all used as suspected crack edges, and the direction consistency degree of two adjacent straight line edges in each suspected crack edge is obtained according to the included angle value of the two adjacent straight line edges in the suspected crack edges; obtaining the regularity of each suspected crack edge in the road edge image according to the direction consistency degree and the flatness of the suspected crack edge; obtaining the final abnormality degree of each linear edge in the road edge image according to the abnormality degree of the linear edge and the regularity of the suspected crack edge where the linear edge is located; and screening crack edges in the road edge image according to the final degree of abnormality, and determining a crack region of the road surface through the crack edges.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the invention, but any modifications, equivalent substitutions, improvements, etc. within the principles of the present invention should be included in the scope of the present invention.