| With the development of the transportation industry,great achievements have been made in the construction of highways in China,and the total length of highways in the country has maintained a growth trend,which has brought great challenges to the maintenance management of pavement.At present,the road surface automatic detection technology gradually replaces the traditional manual detection method,which greatly improves the detection speed and detection accuracy.However,a large amount of road surface detection data cannot be fully and effectively utilized,and the true three-dimensional road surface shape cannot be restored.Therefore,the three-dimensional reconstruction of the pavement and the provision of a more intuitive threedimensional pavement model for the highway maintenance department are the development direction of the pavement detection technology,which is of great significance for the realization of the three-dimensional visualization of the pavement.Firstly,a road surface reconstruction coordinate system is established,and a road surface three-dimensional reconstruction model is obtained,and the principle of the road surface threedimensional reconstruction technology based on the road surface elevation data is proposed.Then,based on the principle of reconstruction,the article carried out the overall scheme design of the road surface 3D reconstruction system,developed the vehicle-mounted road surface 3D reconstruction system,and completed the system hardware selection and software development.An image processing algorithm suitable for laser stripe centerline extraction is proposed in this thesis.The median filter is used to preprocess the acquired laser image and extract the ROI area.Then,the subpixel edge extraction method is used to obtain the upper and lower edge lines of the laser bar,and finally the area between the lines uses the local gray center of gravity method to obtain the precise pixel coordinates of the center line of the laser stripes.A calibration scheme suitable for structured light systems is designed.The relationship between image coordinates and object coordinates is determined by the hidden parameter calibration method.The accuracy of the measurement results is analyzed to verify the feasibility of the system.In order to effectively reduce the frustration of reconstructed pavement,this paper proposes a method of interpolating and fitting the pavement elevation data based on the crosssection profile of the pavement.Catmull-Rom interpolation is used to obtain the smooth crosssection profile.Based on this,a non-uniform rational B-spline surface fitting method is adopted to obtain the road elevation data.At the same time,the influence of vehicle vibration on the system elevation measurement is analyzed,and the road elevation data is corrected.Finally,based on the Microsoft Direct3 D platform,a three-dimensional road surface grid is constructed based on the road surface elevation data,and the road surface image is texturemapped on the vertex grid to obtain a realistic visualized three-dimensional road surface.Through the field test of the reconstruction system,the system measured values of typical road elevation data are compared with the manual measured values to verify that the reconstructed system in this paper has good stability,and the measurement accuracy and repeatability meet the system requirements. |
