CN105627938A

CN105627938A - Pavement asphalt thickness detection method based on vehicle-mounted laser scanning spot cloud

Info

Publication number: CN105627938A
Application number: CN201610008802.6A
Authority: CN
Inventors: 贾宏; 李军; 贾福凯
Original assignee: Xiamen University
Current assignee: Xiamen University
Priority date: 2016-01-07
Filing date: 2016-01-07
Publication date: 2016-06-01

Abstract

The invention discloses a pavement asphalt thickness detection method based on vehicle-mounted laser scanning spot cloud, comprising steps of obtaining spot cloud data, performing data rough registration through choosing a fixed object, using a pavement extraction algorithm based on the road shoulder to respectively detect two sets of the pavement points of the point cloud data, using the road shoulder point to perform precise registration on the data in the block, performing plane structure fitting on the pavement inside the grid, and detecting the asphalt thickness. The invention can fast and accurately detect and modify the thickness of the road asphalt, greatly reduces the processing time and the labor cost, effectively guarantees the road reconstruction quality and provides important data support to the civil traffic safety.

Description

A kind of pavement asphalt thickness detecting method based on Vehicle-borne Laser Scanning point cloud

Technical field

The present invention relates to intelligent transportation system and town road improvement project field, particularly relate to a kind of based on pavement asphalt thickness detecting method in the road reformation engineering of Vehicle-borne Laser Scanning point cloud.

Background technology

Pavement asphalt thickness is by an important indicator of Pavement Structure Design and construction quality inspection. The difference of pavement of road Colophonium thickness can bring the out-of-flatness phenomenon on road surface, and modal is exactly that bituminous paving produces the phenomenons such as crack, rut, wave, hole be recessed. Vehicle travels on the road of poor flatness, and not only fuel consumption is big, comfort level is poor, also can bring certain infringement to vehicle, increases the maintenance cost of vehicle, shortens vehicle service life, even brings certain potential safety hazard. It addition, the out-of-flatness of high speed bituminous paving also can shorten the service life of road, affect economic and social benefit. Therefore accurate measurement road pavement asphalt thickness, periodic detection road quality situation, prepare for super expressway design in future and maintenance, be the problem being worth causing municipal sector greatly to pay close attention to.

At present, traditional Colophonium Thickness sensitivity algorithm mainly has core drilling method and short-pulse radar method two kinds. Core drilling method by directly drilling through sample the thickness of method measured directly detection Colophonium in road surface, but this method does not simply fail to obtain the distribution situation of the complete Colophonium thickness of whole section of road, and it is huge to the injury of road, wastes time and energy, it it is the method that efficiency is minimum. Short-pulse radar method utilizes the method detection Colophonium thickness of radar pulse, although the Colophonium thickness distribution situation obtaining one section of complete road that the method is faster, but being constrained to the difference of road structure and the detection method of pulse echo, its accuracy of detection still cannot well meet the needs of engineering.

In recent years, along with the fast development of laser scanner technique, the Vehicle-borne Laser Scanning system built on basis that is integrated and that control was synchronized at multisensor, it has also become a kind of quick obtaining means of spatial data. On the one hand, vehicle-mounted mobile laser scanning system can realize direct geo-location, and can obtain rapidly scene on a large scale high density, in high precision, three dimensional point cloud that details is abundant; On the other hand, compared with other platform laser scanning systems, vehicle-mounted mobile laser scanning system can obtain the three dimensional point cloud of the more abundant scene on a large scale of higher dot density, higher data precision, details rapidly with less expense.

But, follow-up storage and intelligent processing method are brought serious challenge by the mass data of its acquisition, and meanwhile, the robustness of algorithm is also had higher requirement by the noise existed in data with blocking.

Summary of the invention

It is an object of the invention to provide a kind of based on pavement asphalt thickness detecting method in the road reformation engineering of Vehicle-borne Laser Scanning point cloud.

For achieving the above object, the present invention is by the following technical solutions:

A kind of pavement asphalt thickness detecting method based on Vehicle-borne Laser Scanning point cloud, comprises the following steps:

S1, utilization are equipped with the naked road surface before the collecting vehicle acquisition municipal administration transformation road black top of mobile laser scanning system and two groups of three dimensional point clouds of the bituminous paving after black top;

S2, choose in two groups of data identical, there is the not moving-target such as the roadside poster of obvious corner feature, building, in order under two groups of cloud datas of rough registration to the same coordinate system;

S3, utilize the road surface point detecting two groups of cloud datas based on the road surface extraction algorithm of curb respectively;

S4, for road surface data block corresponding in two groups of cloud datas, utilize the shoulder line of curb point and the matching extracted that it is carried out essence registration;

S5, to merge after road surface data block, utilize the plane characteristic on two road surfaces in the stochastical sampling each piecemeal of consistency algorithm matching;

S6, the Colophonium thickness calculating each road surface piecemeal and earthwork consumption, and then obtain distribution situation and the earthwork total amount of whole section of transformation road asphalt thickness.

Preferably, described step S2 specifically include following step by step:

S21, choose the point that in 4 groups two groups some clouds, moving-target is not corresponding and change basic point to as standard;

S22, utilize these 4 groups conversion basic points, seven parameters in coordinates computed transfer equation;

S23, by transfer equation by two groups of cloud data rough registration under the same coordinate system.

Preferably, described step S3 specifically include following step by step:

S31, original point cloud data is evenly divided into one group of some cloud mass along the direction of wheelpath;

S32, for each some cloud mass, be syncopated as a some cloud along the direction being perpendicular to wheelpath and cut open sheet, and cut open extraction curb point sheet from a cloud;

S33, all curb points are fitted, obtain shoulder line;

S34, original point cloud data is split along shoulder line, obtain road surface cloud data.

Preferably, described step S5 specifically include following step by step:

S51, by merge after essence registration two groups of cloud data blocks, be divided into the grid of 4X4, the road surface cloud data in each grid done following process.

S52, in initial point cloud, randomly choose three points, directly calculate its corresponding flat equation ax+by+z=d, then calculate some Yun Zhongsuo and a little arrive the distance of this plane, d_i=| ax_i+by_i+z_i-d |;

S53, selected threshold t, if certain point arrives the distance d of this plane_i�� t is considered as then interior point, is otherwise invalid data, and adds up the number N drawing the interior point of this plane;

S54, repeat above step, iteration k time, relatively and choose containing the interior maximum plane of point;

S55, re-start plane fitting according to method of characteristic with maximum interior points, draw final fit Plane equation.

Preferably, described step S6 specifically includes following sub-step:

S61, utilize the plane equation of latter two planar structure of matching, calculate the Colophonium thickness on road surface in each grid.

S62, calculating in each piecemeal the angle between two planar structures in grid, formula is as follows:

θ = \cos^{- 1} (\frac{| | F_{1} F_{2} | |}{| | F_{1} | | | | F_{2} | |})

Wherein, F₁And F₂It it is the normal vector of two planes.

S63, one angle threshold �� of setting_t=5 ��, if �� is < ��_t, it is believed that two planar structures are parallel, then the Earthwork calculation formula of this road surface grid is as follows:

V=lwh

Wherein, v is the earthwork of this road surface grid, and l is the length of this grid, and w is the width of this grid, and h is the thickness of this road surface grid Colophonium.

If S64 were �� >=��_t, it is believed that two road surfaces in grid have uneven plane fitting structure, calculate the earthwork of this road surface grid by equation below:

v = \frac{l w}{4} (h_{1} + h_{2} + h_{3} + h_{4}) .

Wherein, h₁��h₂��h₃��h₄The length of side that respectively in grid, four sides are corresponding.

Used by S65, transformation section, the computing formula of the earthwork total amount of Colophonium is: V=��_i,jv��

After adopting technique scheme, the present invention, compared with background technology, has the advantage that

1, the present invention can quickly, accurately, harmlessly detection transformation pavement of road Colophonium thickness, substantially increase detection time and cost, be effectively ensured road construction safety and quality.

2, by combining the smart registration of the rough registration to two groups of cloud datas and monolithic road surface data block, the accuracy of detection of Colophonium thickness is greatly improved.

What 3, propose a kind of novelty utilizes the method for plane fitting to estimate the earthwork consumption of asphalt roads improvement project.

Accompanying drawing explanation

Fig. 1 is the workflow schematic diagram of the present invention.

Fig. 2 be 4 pairs conversion basic points choose schematic diagram.

Fig. 3 is the result schematic diagram after rough registration.

Fig. 4 is extraction and the fitting result schematic diagram of curb point.

Fig. 5 is the result schematic diagram after essence registration.

Fig. 6 is two parallel planar structures schematic diagrams.

Fig. 7 is two non-parallel planes structural representations.

Detailed description of the invention

In order to make the purpose of the present invention, technical scheme and advantage clearly understand, below in conjunction with drawings and Examples, the present invention is further elaborated. Should be appreciated that specific embodiment described herein is only in order to explain the present invention, is not intended to limit the present invention.

Embodiment

Referring to Fig. 1, the invention discloses a kind of based on pavement asphalt thickness detecting method in the road reformation engineering of Vehicle-borne Laser Scanning point cloud, it comprises the following steps:

S1, two groups of cloud datas acquisition

Vehicle-borne Laser Scanning system is utilized to scan acquisition point cloud data set at twice. First time scanning obtains the road cloud data collection of the exposed surface before road reformation engineering asphalt, and second time scanning obtains the road cloud data collection of the bituminous paving after road reformation engineering asphalt.

S2, data rough registration

Be collected in the different time, two groups of cloud data collection of different scanning engineerings need to be registrated under the same coordinate system both before treatment. Two groups of cloud datas have different road surface points, but for the traffic signboard in roadside, light pole, building and road curb etc., they belong to the not moving-target in two groups of cloud datas. For the distance detected between two road surfaces and then the thickness solving Colophonium, we reach the purpose of the whole cloud data of rough registration by the not moving-target selected and mate in cloud data. This step realizes especially by following steps:

S21, choosing 4 groups of points that moving-target is not corresponding in two groups of cloud datas to as conversion basic point (as shown in Figure 2), these 4 groups choosing of basic point of conversion should meet: 1) chooses the angle point of roadside corner feature significantly not moving-target; 2) by the rectangular distribution, road both sides correspondence respectively chooses two pairs; 3) long enough is wanted at two some intervals of road the same side.

S22, by 4 groups change basic points bring Coordinate Transformation Models X respectively into_T=T+ �� RX_SMiddle calculating changes seven parameters.

S23, utilize Coordinate Conversion seven parameter by two groups of cloud data rough registration under the same coordinate system (as it is shown on figure 3, the (a) and (b) figure in Fig. 3 is (c) figure in original scene a and b, Fig. 3 is the result of (a) figure and (b) figure registration. D 1#, 2#, the 3# in () figure is three control point, corresponding curb, tree and traffic mark board).

S3, road surface point data are extracted

Based on wheelpath data, original point cloud data is carried out road surface segmentation, obtains road surface cloud data, referred to herein as wheelpath data be acquired by the inertial navigation system being integrated in Vehicle-borne Laser Scanning system. This step realizes especially by following steps:

S31, along the direction of wheelpath, original point cloud data being evenly divided into one group of some cloud mass, in the present embodiment, the segmentation of some cloud mass is spaced apart 8m.

S32, for each some cloud mass, be syncopated as a some cloud along the direction being perpendicular to wheelpath and cut open sheet, and cut open extraction curb point sheet from a cloud. According to prophet's experience, the curb of road is typically normal to road surface and arranges and higher than road surface, as long as therefore finding the some cloud of highly sudden change namely to have found curb point. The present embodiment extracts the method for curb point specifically, cut open the sheet intersection location with wheelpath to two-sided search from some cloud, find the some cloud of highly sudden change, it can be used as curb point.

S33, all curb points are fitted, obtain curb line (with reference to shown in Fig. 4).

S34, original point cloud data is split along curb line, obtain road surface cloud data.

S4, data essence registration

Through the extraction of road surface point data, two groups of cloud datas after rough registration are divided into a series of corresponding blocks comprising curb point. Utilizing the curb point in block, send out calculating standard based on seven parametric solutions and change matrix, two corresponding blocks are according to not moving-target-curb accuracy registration together (as shown in Figure 5) the most at last.

S5, planar structure matching

To each road surface cloud data block merged after essence registration, it is averaged the grid block being divided into 4X4. Subsequently, each grid block, utilizing two road surfaces in stochastical sampling consistency algorithm fitted mesh difference scheme is planar structure. This step realizes especially by following steps:

S51, to essence registration after merge each road surface cloud data block, be averaged the grid block being divided into 4X4;

S52, initial point cloud in each grid randomly choose three points, directly calculates its corresponding flat equation ax+by+z=d, then calculate some Yun Zhongsuo and a little arrive the distance of this plane, d_i=| ax_i+by_i+z_i-d |;

S6, Colophonium Thickness sensitivity

Detect the Colophonium thickness on this grid road surface by calculating in each grid the distance between two planar structures after matching, and then utilize the relation between planar structure to calculate the earthwork consumption of this grid Colophonium. This step realizes especially by following steps:

θ = \cos^{- 1} (\frac{| | F_{1} F_{2} | |}{| | F_{1} | | | | F_{2} | |})

Wherein, F₁And F₂It it is the normal vector of two planes.

S63, one angle threshold �� of setting_t=5 ��, if �� is < ��_t, it is believed that two planar structures parallel (as shown in Figure 6), then the Earthwork calculation formula of this road surface grid is as follows:

V=lwh

If S64 were �� >=��_t, it is believed that two road surfaces in grid have uneven plane fitting structure (as shown in Figure 7), calculate the earthwork of this road surface grid by equation below:

v = \frac{l w}{4} (h_{1} + h_{2} + h_{3} + h_{4}) .

The above; being only the present invention preferably detailed description of the invention, but protection scope of the present invention is not limited thereto, any those familiar with the art is in the technical scope that the invention discloses; the change that can readily occur in or replacement, all should be encompassed within protection scope of the present invention. Therefore, protection scope of the present invention should be as the criterion with scope of the claims.

Claims

1. A method for detecting pavement asphalt thickness based on vehicle-mounted laser scanning point cloud, is characterized in that: comprise the following steps:

S1. Use the acquisition vehicle equipped with a mobile laser scanning system to obtain two sets of 3D point cloud data of the bare road surface before asphalt paving and the asphalt pavement after asphalt paving of the municipal reconstruction road;

S2. Select the same immovable objects such as roadside billboards and buildings with obvious inflection point characteristics in the two sets of data to roughly align the two sets of point cloud data into the same coordinate system;

S3. Using the shoulder-based road surface extraction algorithm to detect the road surface points of the two sets of point cloud data respectively;

S4. For the corresponding road surface data blocks in the two sets of point cloud data, use the extracted road shoulder points and the fitted road shoulder lines to perform fine registration;

S5. For the fused road surface data block, use a random sampling consensus algorithm to fit the plane features of the two road surfaces in each block;

S6. Calculate the asphalt thickness and earthwork amount of each road surface block, and then obtain the distribution of the asphalt thickness and the total amount of earthwork for the whole section of the reconstructed road.

2. a kind of road surface asphalt thickness detection method based on vehicle-mounted laser scanning point cloud as claimed in claim 1, is characterized in that: described step S2 specifically comprises the following sub-steps:

S21. Selecting 4 groups of point pairs corresponding to the fixed target in the two groups of point cloud data as conversion base points;

S22. Using the four sets of transformation base points, calculate the seven parameters in the coordinate transformation equation;

S23. Roughly register the two sets of point cloud data in the same coordinate system through the transformation equation.

3. A kind of road surface asphalt thickness detection method based on vehicle-mounted laser scanning point cloud as claimed in claim 2, is characterized in that: described step S3 specifically comprises the following sub-steps:

S31. Evenly divide the original point cloud data into a group of point cloud blocks along the direction of the driving track;

S32. For each point cloud block, segment a point cloud slice along the direction perpendicular to the driving track, and extract road shoulder points from the point cloud slice;

S33. Fitting all roadside points to obtain road shoulder lines;

S34. Segment the original point cloud data along the shoulder line to obtain road surface point cloud data.

4. a kind of road surface asphalt thickness detection method based on vehicle-mounted laser scanning point cloud as claimed in claim 3, is characterized in that: described step S5 specifically comprises the following sub-steps:

S51. Divide the two sets of point cloud data blocks that have been fused and finely aligned into 4×4 grids, and perform the following processing on the road surface point cloud data in each grid.

S52. Randomly select three points in the initial point cloud, directly calculate the corresponding plane equation ax+by+z=d, and then calculate the distance from all points in the point cloud to the plane, d _i =|ax _i +by _i + z _i -d|;

S53. Select a threshold t. If the distance d _i ≤ t from a certain point to the plane is considered as an interior point, otherwise it is invalid data, and the number N of points in this plane is obtained by statistics;

S54. Repeat the above steps, iterate k times, compare and select the plane containing the most interior points;

S55. Perform plane fitting again with the most interior points according to the eigenvalue method, and obtain a final fitting plane equation.

5. a kind of road surface asphalt thickness detection method based on vehicle-mounted laser scanning point cloud as claimed in claim 4, is characterized in that, described step S6 specifically comprises the following sub-steps:

S61. Using the plane equations of the two plane structures after fitting, calculate the asphalt thickness of the road surface in each grid;

S62. Calculate the angle between the two plane structures in the grid in each block, the formula is as follows:

θ θ = = {cos cos}^{- - 11} ((\frac{| | | | {F f}_{11} {F f}_{22} | | | |}{| | | | {F f}_{11} | | | | | | | | {F f}_{22} | | | |}))

Among them, F ₁ and F ₂ are the normal vectors of the two planes;

S63. Set an included angle threshold θ _t =5°. If θ<θ _t , it is considered that the two plane structures are parallel, and the earthwork calculation formula of the pavement grid is as follows:

v=lwh

Wherein, v is the earthwork of the pavement grid, l is the length of the grid, w is the width of the grid, and h is the thickness of the asphalt of the pavement grid;

S64. If _θ≥θt , it is considered that the two road surfaces in the grid have non-parallel plane fitting structures, and the earthwork of the road surface grid is calculated by the following formula:

v v = = \frac{l l w w}{44} (({h h}_{11} + + {h h}_{22} + + {h h}_{33} + + {h h}_{44})) . .

Among them, h ₁ , h ₂ , h ₃ , and h ₄ are the side lengths corresponding to the four sides in the grid respectively;

S65. The formula for calculating the total amount of earthwork of asphalt used in the reconstructed road section is: V=Σ _i,j v.