The Research Of Pavement Crack Detection Based On Image

With the rapid development of highway construction in China,the problem of highway maintenance has been paid more and more attention.Pavement crack is one of the common road diseases,and it is also an important indicator of the road quality assessment.The traditional manual detection method has been unable to meet the needs of actual application at present stage.Therefore,it is urgent to study the crack detection system.This paper mainly introduces the research work of pavement crack detection based on image.Firstly,aiming at the characteristics of pavement crack image,the general process and basic treatment method of pavement crack detection are studied,and the algorithm of image preprocessing and target extraction are analyzed and summarized.Secondly,in terms of pavement crack image detection,a crack detection algorithm based on spatial clustering feature is proposed.The image is divided into non-overlapping image blocks to calculate eigenvector and cluster analysis.In order to solve the problem of the image and K-means algorithm.The hierarchical clustering algorithm is used to get the best initial clustering center.The K-means clustering algorithm is used to classify the images into three regions.The fuzzy reasoning algorithm is used to classify the suspected fracture regions.Experimental results demonstrate that the proposed method can detect pavement cracks more accurately compared with K-means algorithm.Thirdly,in terms of pavement crack feature extraction and classification,According to the difference of the geometric shape of the crack,a new feature space is established to judge the type of pavement crack.Meanwhile,the image is segmented by OTSU and morphological processing,and then the road breakage parameters are calculated,which can provide data support for the later highway maintenance.Finally,this paper further studies the pavement crack detection algorithm in a haze environment.Combining with the local filter and local histogram equalization to clear the image.And then through the algorithm proposed in this paper to realize the accurate detection of pavement crack image and the type of crack.