Research On Detection And Measurement Method Of Pavement Distress Based On Depth Image

Civil aviation is an indispensable part of China's transportation industry.With the rapid development of China's economy,the increase of the service life of the airport and the increase of the number of aircraft taking off and landing at the airport,there are inevitable distress on the airport pavement,which reduce the flatness of the pavement and cause serious flight safety risks.In order to ensure the safety of flight take-off and landing,reduce the risk of accidents,early detection of distress and timely maintenance has important practical significance.The main distresses of pavement surface are crack,and the early manifestations of many pavement surface distresses are also cracks.Therefore,the detection and measurement methods of pavement crack distress based on depth image are studied in depth.The main research contents and work are as follows:1.In order to solve the problem of the double strong interference caused by the violent vibration of the platform and the complex apparent structure of the pavement surface,the pavement crack detection algorithm based on curve fitting is proposed.Firstly,an improved fitting model of the main profile curve based on L2 regularization constraints is designed.Secondly,the improved greedy strategy is used to optimize the fitting model.Finally,the main profile of the multi-direction airfield is extracted and integrated,and the crack depth and shape information are used to extract real distress.The experimental results show that the proposed algorithm can effectively extract the main profile and realize the accurate detection of crack.2.In order to solve the problem that the main profile of a single section has no good continuity in 3d space,the method based on surface model fitting is proposed?Firstly,the depth image is divided into several overlapping grid regions,and local cubic surfaces are constructed for each grid region based on RANSAC algorithm.Then,the surface models of all grid regions are fused to generate the global surface model of the whole image acquisition region.Finally,by using the difference between the global surface model and the original depth image,the candidate crack pixels are segmented by the adaptive threshold method,and the candidate crack pixels are screened by a variety of morphological constraints.The experiment results show that the algorithm increases the three-dimensional spatial continuity in the main profile extraction and obtains better crack detection results.3.For the extracted 3d crack information,a method of measuring crack length and width based on morphology is proposed.Firstly,K3 M algorithm is used to extract the skeleton of crack and calculate the skeleton length to obtain the crack length.Then the width of the crack is obtained by calculating the distance between the upper and lower edges of the crack and the local intersection of the skeleton normal.The proposed algorithm can measure the basic characteristics of the disease and provide technical support for the accurate calculation and repair of the characteristic parameters of the surface distress.