Research On Crack Image Processing Algorithm Of Asphalt Pavement Based On Computer Vision

With the continuous progress of China’s transportation industry,China’s highway industry is gradually transitioning from construction to construction and maintenance.Asphalt pavement is the main structural form of highway pavement in Our country.Pavement crack is the main form of pavement disease.At present,gray scale or color digital image is mainly used for detection,which has the advantages of fast detection speed,but easy to be interfered and low accuracy.Computer vision can make the computer have the ability to identify the depth information in the image.This thesis combines computer vision with traditional digital image detection to overcome the problem of low detection accuracy of traditional detection methods.The thesis analyzes the distribution of asphalt pavement diseases in Bijie city of Guizhou Province and draws a conclusion that the main diseases of asphalt pavement in Bijie are cracks.The characteristics of digital images of asphalt pavement cracks are analyzed,and the collection method of pavement cracks images is studied.The study shows that the overlap rate of adjacent images should be about 70%.After preprocessing the pavement image,SURF algorithm was used to search the feature points,and then feature points were evaluated according to the Euclidean distance of feature points.Finally,the continuous asphalt pavement image was obtained by weighted fusion method.Use of image enhancement,image segmentation,image morphology processing method on the cracks of asphalt pavement,pavement crack skeleton of binary image,and according to the cracks on the X axis and Y axis of projection and the size of the connected domain empty part of the area determine the type of fracture,and to determine the possible fracture area and position,It lays a foundation for the subsequent accurate detection based on computer vision.In order to improve the anti-interference ability of pavement crack recognition,computer vision is introduced.For computer vision are introduced,using binocular computer vision system is proposed to test the crack,to study the binocular vision,the calculation principle of ilf calibration method is used to get the camera parameters,using the distortion correction,polar correction has carried on the correction of image,using a global stereo matching algorithm to generate the depth map,and the depth map to a into the gray image,Identification again,accurate analysis of cracks,can get the location,type,size and depth of cracks information.The software system and hardware platform are designed by using the detection method to analyze the influence of color,shadow and light on the anti-interference results.It is applied to the actual highway detection in Bijie area,Guizhou Province,and is compared with the manual detection method.The results show that the method has high detection accuracy,and the test results are more credible.