Study On The Detection System Of Asphalt Pavement Crack Based On Image Processing

For the entire highway pavement maintenance management system, the asphalt highway pavement cracks automatic detection technology has very important meaning. The traditional method of detection is unable to fit the needs of modern detection. Along with the development of computer technology, the road detection technology is combined with computer technology, and enters the automatic detection age. The author participated in the research project of the Key Laboratory of the Ministry of Education - the image detection, identification and automatic evaluation methods research of asphalt pavement cracks, and studied highway asphalt pavement cracks image detection system.The main content of this thesis summarizes as follows:1. Proposes modified median filtering algorithm. The original enhancement algorithm may easily fuzz edges of cracks while smoothing images. However, it removes most of the noise, and maintains edges well via this advanced algorithms, thus provides high-quality digital images for the edge detection.2. Applies improved Canny algorithm to detect edges in the crack images, as the classical edge detection algorithm runs poor, and prone to form pseudo-edge. The thesis applies an improved threshold selection based Canny algorithm to detect crack images. The experiment achieves satisfactory.3. Designs a SQL database system, and establishes the SQL database application system which fits all kinds of Highway Maintenance Management System, based on the practical application environment and business operation processes of the highway Maintenance Management System.4. Designs a highway road detection system. The thesis designs and tests the system modularly, according to the characteristics of highway detection system. The experimental outcome is analyzed, and a brief statement about the existent problems is given.