Research On Pavement Defect Detection And Point Cloud Repair Technolngy Based On Image And Point Cloud

With the development of China’s social economy,highway mileage is also expanding,the road damage not only seriously affects the road service level but also endangers driving safety,therefore,the detection of road defect has become the main problem that traffic management departments pay attenti on to.The traditional manual detection of pavement defects is not only inefficient,but also has great security risks.In this paper,laser scanner and linear array camera are used to scan the pavement in real time.By collecting pavement point cloud data and image data,defect detection is carried out to improve the detection efficiency and accuracy of pavement defects.Repair point cloud road surface defects at the same time,its purpose from the Angle of the computer simulation was carried out on the ro ad surface defect repair,through the defect point cloud to repair as a result,the analysis of the road roughness index,which reflects the real distribution of point cloud,zero defect pavement in order to satisfy the demands of pavement maintenance work,and to provide the actual road surface defect fix reasonable and efficient technical support.Firstly,this paper studies the collection method of common pavement defects,classifies and grades pavement defects according to "Highway Technical Condition Assessment Standard",and selects the appropriate method to complete the data collection of pavement defects in combination with the characteristics of actual pavement defects.Then the detection of pavement defects based on image and point cloud is studied.The detection method of image features is adopted for pavement crack defects.By preprocessing the collected crack images,the image noise is effectively removed,and the grayscale details of crack area features are highlighted,so that the pavement crack images are enhanced.By combining the advantages of OTSU and Canny edge detection algorithm,the pavement crack image is segmented,and the complete crack binary image is obtained.Finally,the geometric characteristics of pavement crack are analyzed and calculated.For pavement pit slot and upheaval defects point cloud detection method and by means of surface defect preprocessing of point cloud to achieve effective to streamline the characteristics of regional point cloud,and remove the device or the external environment factors such as the impact of outliers,through the defect of the point cloud height information,intuitive response surface concave and convex,by building surface analysis of the curvature is negative,It can also realize the detection of road potholes and upheaval.Finally,the repair method based on the pavement defect point cloud was studied.Aiming at the pavement crack point cloud,the mapping relationship between the image and the point cloud was established through coordinate transformation,and the radial basis function was used to construct the implicit surface method to realize the crack point cloud repair.Aiming at the road pothole and upheaval point clouds,the road pothole and upheaval point clouds were repaired layer by layer by using the minimum outer rectangle of the point cloud in the defect area and the mean elevation of the adjacent area of edge points.In order to verify the effectiveness of the method used in this paper,a large number of experiments were carried out in the detection of pavement defects based on image and point cloud,and the detection results could all reach the defect identification rate of over 90%,achieving the expected detection effect.In the pavement defect repair experiment based on point cloud,the repaired pavement point cloud and the point cloud at the defect edge connect and transition naturally,and meet the requirements of the pavement smoothness index.Further research will be carried out to improve defect detection accuracy and image b ased pavement defect repair.