A method for road surface point cloud smoothing and crack information processing based on moving window

Abstract

The present invention discloses a method for smoothing and crack information processing of pavement point clouds based on a moving window, and belongs to the technical field of pavement crack disease detection. The method comprises: S10, collecting original three-dimensional point cloud data of the target area, and processing the collected data in blocks to obtain a pavement depth matrix, and applying a neighborhood mean-variance threshold to remove discrete points; S20, according to the pavement depth matrix processed in step S10, using the median of each window obtained by a moving window to fit the real surface of the pavement, and constructing a potential crack point cloud set for the pavement depth matrix point cloud below the pavement surface; S30, based on the potential crack point cloud set, using the longitudinal and transverse depth mappings in the longitudinal and transverse directions of the road to obtain the real crack area, and determine the crack start and end position information and its maximum depth. The method of the present invention greatly improves the performance and efficiency of the three-dimensional point cloud crack recognition algorithm.

Description

Technical Field

[0001] The present invention belongs to the technical field of pavement crack disease detection, and more specifically, relates to a pavement point cloud smoothing and crack information processing method based on a moving window. Background Art

[0002] Asphalt pavement, as an important part of highway engineering, is an important infrastructure to promote economic development. With the increase of years of use, asphalt pavement has developed various forms of diseases, among which pavement cracks have become the most important disease of highways and urban roads.

[0003] The manual inspection and judgment method has problems such as low work efficiency, high labor intensity, poor reproducibility, high labor cost, and low accuracy. In recent years, a large number of studies have used digital image processing technology to identify asphalt pavement defects. However, the image signal-to-noise ratio is relatively low, and the target information cannot be accurately extracted, and the application effect is not satisfactory. The three-dimensional point cloud device can obtain millimeter-level elevation information of the road surface, and can avoid the influence of adverse conditions such as lighting conditions, shadows, and oil pollution, providing a new research direction for efficient and accurate crack detection technology.

[0004] At present, the common pavement 3D point cloud crack information extraction mainly includes the following:

[0005] (1) Based on the crack information extraction method converted into a two-dimensional grayscale image, the point cloud data is converted into depth data, and then further converted into a grayscale image. The crack information is extracted through the crack extraction method of the two-dimensional image. Compared with traditional image processing, this method improves the accuracy of the original data, but cannot eliminate the interference of texture and noise;

[0006] (2) Crack information extraction method based on mathematical morphology, using height difference, normal vector, gradient, second-order derivative and other data to extract possible crack points that meet the mathematical characteristics of cracks, and generate crack skeletons through morphological methods. Due to the differences in pavement materials, environment, and traffic volume, this method cannot be applied to all crack point clouds;

[0007] (3) The crack information extraction method based on deep learning and neural network uses the relevant technologies of neural network to build a network based on the characteristics of cracks and point clouds. However, this method requires manual labeling at the pixel level for training, which requires a certain amount of time and labor costs.

[0008] Under the existing methods, problems such as low crack identification accuracy and low level of identification intelligence still exist. Summary of the invention

[0009] In order to solve at least one of the above technical problems, according to one aspect of the present invention, a method for smoothing road point cloud and processing crack information based on a moving window is provided, comprising the following steps:

[0010] S10, collecting original three-dimensional point cloud data of the target area, and processing the collected data in blocks to obtain a road surface depth matrix, and applying a neighborhood mean-variance threshold to remove discrete points;

[0011] S20, according to the road depth matrix processed in step S10, the median of each window obtained by moving the window is used to fit the real surface of the road surface, and a potential crack point cloud set is constructed for the road depth matrix point cloud below the road surface;

[0012] S30. Based on the potential crack point cloud set, the real crack area is obtained by using the longitudinal and transverse depth mapping in the longitudinal direction and transverse direction of the road, and the crack start and end point position information and the maximum depth are determined.

[0013] According to the moving window-based pavement point cloud smoothing and crack information processing method of the embodiment of the present invention, optionally, the step S10 specifically includes:

[0014] S11, converting the discrete point cloud set Ω with unequal spacing into a road surface depth matrix Z according to a set resolution S;

[0015] Ω={p1,p2,…p N}, p i ={x i ,y i ,z i}∈R 3 ;

[0016]

[0017] Where, the horizontal coordinate X={x1,x2…,x i ,…,x m}, x1 <x2<…<x i <… <x m , x i+1 -x i =S, vertical coordinate Y = {y1,y2,…,y j ,…,y n},y1 <y2<…<y j <… <y n ,y j+1 -y j =S,

[0018] S12, slicing the road depth matrix Z in step S11 along the X direction and the Y direction to construct N road depth matrices of size k×k and resolution S, where N is the total number of road depth matrices after segmentation, N=(m / / k)×(n / / k);

[0019] S13. Apply the neighborhood mean-variance threshold to N road depth matrices of size k×k and resolution S, perform a statistical analysis on the neighborhood of each point, calculate its average distance to all nearby points, remove the point clouds with average distances outside the standard range from the data, and obtain the processed road depth matrix Z′.

[0020] According to the moving window-based pavement point cloud smoothing and crack information processing method of the embodiment of the present invention, optionally, the step S20 specifically includes:

[0021] S21, according to any processed road surface depth matrix Z′, select a moving window k′×k′ whose area is larger than the crack area, take the median of the depth of each window and regard it as the real undamaged road surface depth;

[0022] S22, repeating step S21 until all processed road depth matrix Z′ values ​​are traversed, fitting all points through cubic spline interpolation to obtain a smooth road surface Γ, and calculating the road real depth matrix Z″;

[0023] S23. Subtract the depth value of the road surface real depth matrix Z″ from the depth value of the processed road surface depth matrix Z′ to obtain the depth of the point cloud relative to the road surface, thereby constructing a potential crack point cloud set q.

[0024] According to the moving window-based road point cloud smoothing and crack information processing method according to an embodiment of the present invention, optionally, step S30 specifically includes:

[0025] S31, map the depth value of the point set q generated in step S23 to the xoz plane to obtain a discrete point set d X ={d X1 ,d X2 ,…,d Xm}; Map the depth values ​​of the point set q generated in step S23 to the yoz plane to obtain a discrete point set d Y ={d Y1 ,d Y2 ,…,d Yn};

[0026] S32, for discrete point set d X and d Y , construct the circumscribed rectangle T = {(x,y)|x min ≤x≤x max ,y min ≤y≤y max}, obtain the corresponding vertex position information (x min ,x max ,y min ,y max );

[0027] S33, according to the circumscribed rectangle T = {(x, y) | x min ≤x≤x max ,y min ≤y≤y max}, extract the maximum depth value d in the area max , that is, the maximum depth z min .

[0028] According to the moving window-based pavement point cloud smoothing and crack information processing method of an embodiment of the present invention, optionally, in step S10, if the pavement depth matrix data obtained by segmenting the three-dimensional point cloud data is missing by more than 20%, it is considered that the collection is erroneous and the segmentation result is deleted.

[0029] According to the moving window-based pavement point cloud smoothing and crack information processing method of the embodiment of the present invention, optionally, in step S20, 2D≤k′≤k / 20, and D is the normal crack width at the set resolution.

[0030] According to the moving window-based pavement point cloud smoothing and crack information processing method of the embodiment of the present invention, optionally, in step S30, for the discrete point set d X and d Y , when z X From 0 to positive, and from positive to 0, it is considered that the corresponding starting and ending abscissas of the crack are x min 、x max , when z Y From 0 to positive, and from positive to 0, it is considered that the starting and ending ordinates of the crack are y min ,y max .

[0031] According to the moving window-based road point cloud smoothing and crack information processing method of the embodiment of the present invention, optionally, in step S30, the discrete point set d X and d Y When the length of continuous 0 values ​​exceeds 0.05 m, the crack is considered to be naturally interrupted and there are multiple linear cracks.

[0032] According to the moving window-based pavement point cloud smoothing and crack information processing method of an embodiment of the present invention, optionally, for a plurality of linear cracks, crack circumscribed rectangle parameters and maximum depth are extracted for each crack according to steps S32 and S33.

[0033] According to another aspect of the present invention, a computer-readable storage medium is provided, and when the program is executed by a processor, the program implements the steps of the moving window-based road surface point cloud smoothing and crack information processing method of the present invention.

[0034] Beneficial Effects

[0035] Compared with the prior art, the present invention has at least the following beneficial effects:

[0036] The moving window-based pavement point cloud smoothing and crack information processing method of the present invention fits the real pavement surface through the moving window method, can be used for rapid identification of pavement crack position and morphological information, and can also be used for calculation of the degree of damage of multiple intersecting pavement linear cracks. It is scientific, reasonable and easy to implement. Its promotion and application are of great significance to the application of pavement point cloud information and rapid crack determination, and greatly improves the performance and efficiency of the three-dimensional point cloud crack recognition algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings of the embodiments will be briefly introduced below. Obviously, the drawings in the following description only relate to some embodiments of the present invention, but are not intended to limit the present invention.

[0038] Figure 1 The flow chart of the method for smoothing road surface point cloud and crack information based on moving window of the present invention is shown;

[0039] Figure 2 A schematic diagram of a moving window in Embodiment 1 of the present invention is shown;

[0040] Figure 3 One of the block point cloud data after removing the discrete points in Example 2 of the present invention is shown;

[0041] Figure 4 The road surface fitted by the moving window depth median in Example 2 of the present invention is shown;

[0042] Figure 5 The potential crack point set extracted in Example 2 of the present invention is shown;

[0043] Figure 6 The lateral depth projection of the road in Embodiment 2 of the present invention is shown;

[0044] Figure 7 The longitudinal depth projection of the road in Embodiment 2 of the present invention is shown;

[0045] Figure 8 The longitudinal depth projection of the road in the comparative example is shown;

[0046] Fig. 9 The lateral depth projection of the road in the comparative example is shown. DETAILED DESCRIPTION

[0047] To make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the embodiments of the present invention will be described clearly and completely in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them.

[0048] Unless otherwise defined, technical or scientific terms used herein shall have the same meaning as commonly understood by one of ordinary skill in the art to which the invention belongs.

[0049] In order to reduce the interference to the road texture and improve the application scope of the crack extraction method, the present invention fits the road surface based on the moving window method, combines depth mapping, obtains the position and morphological information of the crack point cloud, and realizes accurate identification of road crack information. It has the technical advantages of fast, intelligent and efficient, and can be widely used in the field of road crack disease detection technology.

[0050] Example 1

[0051] The pavement point cloud smoothing and crack information processing method based on the moving window of this embodiment is as follows: Figure 1 As shown, the following steps are included.

[0052] S10, collecting the original three-dimensional point cloud data of the target area, and processing the collected data in blocks to obtain a road depth matrix, and applying a neighborhood mean-variance threshold to remove discrete points, specifically including:

[0053] S11, converting the discrete point cloud set Ω with unequal spacing into a road surface depth matrix Z according to a set resolution S;

[0054] Ω={p1,p2,…p N}, p i ={x i ,y i ,z i}∈R 3 ;

[0055]

[0056] Where, the horizontal coordinate X={x1,x2…,x i ,…,x m}, x1 <x2<…<x i <… <x m , x i+1 -x i =S, vertical coordinate Y = {y1,y2,…,y j ,…,y n},y1 <y2<…<y j <… <y n ,y j+1 -y j=S,

[0057] R 3 is a set of rational numbers in three-dimensional space, m is the number of image rows, and n is the number of image columns;

[0058] S12, slicing the road depth matrix Z in step S11 along the X direction and the Y direction to construct N road depth matrices of size k×k and resolution S, where N is the total number of road depth matrices after segmentation, N=(m / / k)×(n / / k);

[0059] S13. Apply the neighborhood mean-variance threshold to X road depth matrices of size k×k and resolution S, perform a statistical analysis on the neighborhood of each point, calculate the average distance from it to all adjacent points, remove the point clouds with average distances outside the standard range from the data, and obtain the processed road depth matrix Z′.

[0060] The meaning of the standard range described in this embodiment is as follows: assuming that the average distance from each point to all adjacent points is a Gaussian distribution, whose shape is determined by the mean ε and the standard deviation σ. The standard range is defined as [ε-σ,ε+σ]. Points with an average distance outside the standard range are outliers. Outliers are mostly abnormal data and need to be eliminated to avoid affecting the accuracy of the processing results.

[0061] Furthermore, there may be data loss when the device collects original 3D point cloud data. If the road depth matrix data obtained by segmenting the 3D point cloud data is missing by more than 20%, the collection is considered to be incorrect and the segmentation result is deleted.

[0062] S20, according to the road depth matrix processed in step S10, the median of each window obtained by moving the window is used to fit the real surface of the road surface, and a potential crack point cloud set is constructed for the road depth matrix point cloud below the road surface, specifically including:

[0063] S21, according to any processed road surface depth matrix Z′, select a moving window k′×k′ whose area is larger than the crack area, take the median of the depth of each window and regard it as the real undamaged road surface depth;

[0064] The moving window size is 2D≤k′≤k / 20. If the window size k′ is set too large, the fitted road surface will lose some texture details. Otherwise, more crack points will be misjudged as real road surfaces.

[0065] D is the normal crack width at the set resolution;

[0066] like Figure 2As shown in the figure, the black is the normal point cloud depth of the road surface, the gray is the crack point, and the minimum value of k′ is selected as twice the normal crack width at the set resolution. For example, when the resolution S = 1mm, the crack width is 5mm, that is, the width of 5 pixels, so the minimum value of k′ is 10; in addition, when the resolution S = 10mm, the crack width is 5mm, that is, the width of 0.5 pixels, but when k′ is 1 or 2, taking the median is likely to cause errors, so the minimum k′ cannot be less than 3; the maximum value of k′ is k / 20, which ensures that the number of points in the point cloud that fit the normal plane is sufficient and does not lose the local road surface depth changes affected by rutting.

[0067] S22, repeating step S21 until all processed road depth matrix Z′ values ​​are traversed, fitting all points through cubic spline interpolation to obtain a smooth road surface Г, and calculating the road real depth matrix Z″;

[0068] S23. Subtract the depth value of the road surface real depth matrix Z″ from the depth value of the processed road surface depth matrix Z′ to obtain the depth of the point cloud relative to the road surface, thereby constructing a potential crack point cloud set q.

[0069] The reason why the median depth is used to fit the road surface in step S20 of this embodiment is that the number of crack points is less than that of normal road surface points. Therefore, the median depth of each moving window can be considered as the depth of normal road surface points. The road surface fitted by the median is considered to be the road point cloud depth in the absence of cracks. Finally, the original point cloud data and the fitted road surface are compared. The depth lower than the fitted original point cloud depth is considered to be a potential crack point set.

[0070] S30, based on the potential crack point cloud set, using the longitudinal and transverse depth mapping in the longitudinal direction and transverse direction of the road, obtain the real crack area, determine the crack start and end point position information and the maximum depth, specifically including:

[0071] S31, map the depth value of the point set q generated in step S23 to the xoz plane to obtain a discrete point set d X ={d X1 ,d X2 ,…,d Xm}; Map the depth values ​​of the point set q generated in step S23 to the yoz plane to obtain a discrete point set d Y ={d Y1 ,d Y2 ,…,d Yn};

[0072] S32, for discrete point set d X and d Y , construct the circumscribed rectangle T = {(x,y)|x min ≤x≤x max ,ymin ≤y≤y max}, obtain the corresponding vertex position information (x min ,x max ,y min ,y max );

[0073] S33, according to the circumscribed rectangle T = {(x, y) | x min ≤x≤x max ,y min ≤y≤y max}, extract the maximum depth value d in the area max , that is, the maximum depth z min .

[0074] Furthermore, for the discrete point set d X and d Y , when z X From 0 to positive, and from positive to 0, it is considered that the corresponding starting and ending abscissas of the crack are x min 、x max , when z Y From 0 to positive, and from positive to 0, it is considered that the starting and ending ordinates of the crack are y min ,y max ; Discrete point set d X and d Y When the length of continuous 0 values ​​exceeds 0.05 m, it is considered that the crack is naturally interrupted and there are multiple linear cracks. For multiple linear cracks, the crack circumscribed rectangle parameters and the maximum depth are extracted for each crack according to steps S32 and S33.

[0075] Example 2

[0076] The pavement point cloud smoothing and crack information processing method based on the moving window of this embodiment is as follows:

[0077] For the original point cloud data collected, the discrete point cloud set Ω={p1,p2,…p N}, p i ={x i ,y i ,z i}∈R 3 Calculate the mean depth according to the set resolution S = 1mm to achieve the conversion of the road depth matrix Z:

[0078]

[0079] The horizontal coordinate X={x1,x2…,x i ,…,x m}, x1 <x2<…<x i <… <xm , x i+1 -x i =S, vertical coordinate Y = {y1,y2,…,y j ,…,y n},y1 <y2<…<y j <… <y n ,y j+1 -y j =S,

[0080] Taking into account the accuracy of existing equipment and the accuracy required for crack extraction, it is recommended that S=1mm~10mm. In this embodiment, S=1mm.

[0081] The obtained road depth matrix Z is divided into blocks along the X direction and the Y direction to construct N road depth matrices of k×k size and resolution S, where N is the total number of road depth matrices after blocking, N=(m / / k)×(n / / k). To ensure the processing efficiency of the depth matrix, it is recommended that k≤2048. In this embodiment, k=600.

[0082] Take one of the block results for further explanation.

[0083] like Figure 3 As shown, the neighborhood mean-variance threshold is used to remove the discrete points of the road depth matrix block data with a size of k×k and a resolution of S to obtain the road depth matrix block Z′.

[0084] like Figure 4 As shown in the figure, according to the processed road depth matrix block Z′, the real road surface plane Г is fitted by the moving window method, as shown in Figure 5 As shown, the point cloud below the road surface is obtained to construct a potential crack point cloud set q, where k′=15;

[0085] The generated crack point cloud set q is mapped using the depth in the longitudinal and transverse directions of the road, as Figure 6 As shown, projecting onto the xoz plane, we get the discrete point set d X ={d X1 ,d X2 ,…,d Xm}, m is the number of discrete points corresponding to the crack length; Figure 7 As shown, projecting onto the yoz plane, we get the discrete point set d Y ={d Y1 ,d Y2 ,…,d Yn}, n is the number of discrete points corresponding to the crack width.

[0086] For the discrete point set d X and d Y , when ZX From 0 to positive, and from positive to 0, it is considered that the corresponding starting and ending abscissas of the crack are x min 、x max , when z Y From 0 to positive, and from positive to 0, it is considered that the starting and ending ordinates of the crack are y min ,y max , construct the circumscribed rectangle T = {(x,y)|x min ≤x≤x max ,y min ≤y≤y max}, get the real crack area and its corresponding vertex position information (376,541,0,600).

[0087] According to the circumscribed rectangle T = {(x,y)|x min ≤x≤x max ,y min ≤y≤y max}, by obtaining the maximum depth value d in the area max , determine the maximum crack depth z min =7.422mm.

[0088] Comparative Example

[0089] The method steps S10 and S30 used in the comparative example and the embodiment 2 are the same, except that the comparative example step S20 adopts the Chinese patent publication number CN113256805A, a method for fast calculation of road surface linear crack information based on three-dimensional point cloud reconstruction, to construct a potential linear crack point cloud set. By the comparative method, the longitudinal depth projection of the road is obtained as follows: Figure 8 As shown, the road lateral depth projection is as follows Fig. 9 As shown, Figure 8 The maximum depth change in the longitudinal direction of the road was calculated. The depth of the road point cloud dropped significantly between 300 pixels and 400 pixels. Fig. 9 It shows the changes of cracks in the horizontal direction of the road. Between 50 pixels and 150 pixels, the road point cloud depth is concave compared with the overall linear trend. However, the point cloud depth is affected by the change of the surrounding surface height. Therefore, the specific location of the crack can only be analyzed by combining the point cloud depth change trend. Figure 6 and Figure 7 Compared with the comparative Figure 8 and Fig. 9 ,The maximum depth of the original point cloud before moving window processing includes not only the influence of cracks, but also the influence of road slope inclination. Moving window processing is a necessary means for road point cloud data with slope, depression and ridge. After moving window processing, the influence of cracks on depth is more significant, such as Figure 7The cracks are located between 400 and 500 pixels, and depth fluctuations occur in the transverse direction of the road, proving that the linear crack runs through the transverse direction of the road. The range of variation in the point cloud depth is also reduced to 8 mm, which is more in line with the actual situation.

[0090] Example 3

[0091] The computer-readable storage medium of this embodiment stores a computer program thereon, and when the program is executed by a processor, the steps in the method for smoothing road point cloud and crack information based on a moving window of Embodiment 1 are implemented.

[0092] The computer-readable storage medium of this embodiment may be an internal storage unit of the terminal, such as a hard disk or memory of the terminal; the computer-readable storage medium of this embodiment may also be an external storage device of the terminal, such as a plug-in hard disk, a smart memory card, a secure digital card, a flash memory card, etc. equipped on the terminal; further, the computer-readable storage medium may also include both an internal storage unit of the terminal and an external storage device.

[0093] The computer-readable storage medium of this embodiment is used to store computer programs and other programs and data required by the terminal. The computer-readable storage medium can also be used to temporarily store data that has been output or is to be output.

[0094] Example 4

[0095] The computer device of this embodiment includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps in the moving window-based pavement point cloud smoothing and crack information processing method of Embodiment 1 are implemented.

[0096] In this embodiment, the processor may be a central processing unit, or other general-purpose processors, digital signal processors, application-specific integrated circuits, readily available programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include a non-volatile random access memory. For example, the memory may also store information about the device type.

[0097] Those skilled in the art will appreciate that the disclosed content of the embodiments may be provided as methods, systems, or computer program products. Therefore, the present solution may adopt the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Moreover, the present solution may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program codes.

[0098] The present solution is described with reference to the flowchart and / or block diagram of the method and computer program product according to the embodiment of the present solution. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions; these computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0099] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0101] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0102] The examples described in the present invention are merely descriptions of the preferred implementation modes of the present invention, and are not intended to limit the concept and scope of the present invention. Without departing from the design concept of the present invention, various modifications and improvements made to the technical solutions of the present invention by engineers and technicians in this field should all fall within the protection scope of the present invention.

Claims

1. A method for smoothing road point cloud and processing crack information based on a moving window, characterized in that: The steps include: S10, collecting original three-dimensional point cloud data of the target area, and processing the collected data in blocks to obtain a road surface depth matrix, and applying a neighborhood mean-variance threshold to remove discrete points; S20, according to the road depth matrix processed in step S10, the median of each window obtained by moving the window is used to fit the real surface of the road surface, and a potential crack point cloud set is constructed for the road depth matrix point cloud below the road surface; S30, based on the potential crack point cloud set, using the longitudinal and transverse depth mappings in the longitudinal direction and transverse direction of the road, obtain the real crack area, determine the crack start and end point position information and the maximum depth; The step S10 specifically includes: S11, converting the discrete point cloud set Ω with unequal spacing into a road surface depth matrix Z according to a set resolution S; Ω={p1,p2,…p N },p i ={x i ,y i ,from i }∈R 3 ; Where, the horizontal coordinate X={x1,x2…,x i ,…,x m }, x1 <x2<…<x i <… <x m , x i+1 -x i =S, vertical coordinate Y = {y1,y2,…,y j ,…,y n },y1 <y2<…<y j <… <y n ,y j+1 -y j =S, S12, slicing the road depth matrix Z in step S11 along the X direction and the Y direction to construct N road depth matrices of size k×k and resolution S, where N is the total number of road depth matrices after segmentation, N=(m / / k)×(n / / k); S13, applying the neighborhood mean-variance threshold to N k×k-sized, resolution S road depth matrices, performing a statistical analysis on the neighborhood of each point, calculating the average distance from it to all adjacent points, removing the point clouds with average distances outside the standard range from the data, and obtaining the processed road depth matrix Z'; The step S20 specifically includes: S21, according to any processed road surface depth matrix Z', select a moving window k'×k' with an area larger than the crack area, take the median depth of each window and regard it as the real undamaged road surface depth; S22, repeating step S21 until all processed road depth matrix Z' values ​​are traversed, fitting all points through cubic spline interpolation to obtain a smooth road surface Γ, and calculating the road real depth matrix Z"; S23, subtracting the depth value of the road surface real depth matrix Z" from the depth value of the processed road surface depth matrix Z' to obtain the depth of the point cloud relative to the road surface, thereby constructing a potential crack point cloud set q; The step S30 specifically includes: S31, map the depth value of the point set q generated in step S23 to the xoz plane to obtain a discrete point set d X ={d X1 ,d X2 ,…,d Xm }; Map the depth values ​​of the point set q generated in step S23 to the yoz plane to obtain a discrete point set d Y ={d Y1 ,d Y2 ,…,d Yn }; S32, for discrete point set d X and d Y , construct the circumscribed rectangle T = {(x,y)|x min ≤x≤x max ,y min ≤y≤y max }, obtain the corresponding vertex position information (x min ,x max ,y min ,y max ); S33, according to the circumscribed rectangle T = {(x, y) | x min ≤x≤x max ,y min ≤y≤y max }, extract the maximum depth value d in the area max , that is, the maximum depth z min .

2. The method for smoothing road point cloud and processing crack information based on a moving window according to claim 1, characterized in that: In the step S10, if the road depth matrix data obtained by dividing the three-dimensional point cloud data into blocks is missing more than 20%, it is considered that the collection is wrong and the block result is deleted.

3. The method for smoothing road point cloud and processing crack information based on a moving window according to claim 1, characterized in that: In the step S20, 2D≤k'≤k / 20, and D is a normal crack width at a set resolution.

4. The method for smoothing road point cloud and processing crack information based on a moving window according to claim 1, characterized in that: In step S30, for the discrete point set d X and d Y , when z X From 0 to positive, and from positive to 0, it is considered that the corresponding starting and ending abscissas of the crack are x min 、x max , when z Y From 0 to positive, and from positive to 0, it is considered that the starting and ending ordinates of the crack are y min ,y max .

5. The method for smoothing road point cloud and processing crack information based on a moving window according to claim 1, characterized in that: In step S30, the discrete point set d X and d Y When the length of continuous 0 values ​​exceeds 0.05 m, the crack is considered to be naturally interrupted and there are multiple linear cracks.

6. The method for smoothing road point cloud and processing crack information based on a moving window according to claim 5, characterized in that: For multiple linear cracks, the crack circumscribed rectangle parameters and the maximum depth are extracted for each crack according to step S32 and step S33.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the method for smoothing road point cloud and crack information based on a moving window as described in any one of claims 1 to 6 are implemented.

Images