Study On The Key Technology Of Asphalt Pavement Cracks Automatic Detection And Repairment

In this paper,the real road image collected by digital camera is used as the data basis.Firstly,image preprocessing was carried out for the problems of noise,shadow and low contrast.Secondly,the pre-processed image is further processed to extract road crack information in the image;finally,the road crack automatic repair path is planned based on the road crack information extracted from the image,and a path tracking control scheme is proposed.The main research results of this paper are:(1)Image preprocessing methods: After experimenting and comparing many image preprocessing methods,it is found that the average method is used to grayscale the image,the method based on partial differential equation is used to reduce the noise,and the image is segmented.The linear enhancement algorithm can enhance the contrast of the image and get better results.(2)Pavement crack skeleton extraction method: Aiming at the problem that the existing algorithm has poor adaptability in different scenarios and inaccurate detection results,a tensor-based voting algorithm is proposed.Method of road surface crack skeleton extraction,and the method is optimized to improve the detection speed.An evaluation method of crack detection effect was proposed,and the detection effect of this paper was evaluated.The results show that the accuracy rate of extracting the crack information of the road surface using the tensor voting algorithm can reach more than 90%.(3)Path planning and path tracking control scheme: Established the conversion relationship from image coordinates to real-world coordinates,proposed a method for planning crack repair paths based on the extracted road crack skeleton information,and designed a fuzzy adaptive PID path tracking control system.The simulation experiment of the path tracking control system is carried out through the Matlab Simulink environment.The experiments show that the fuzzy adaptive PID path tracking control system designed in this paper can well control the robotic arm to complete the joint filling task.