Automatic Detection System Of Road Surface Disease Architectural Improvements And The Detection Algorithm

This paper is written on the background of the project of N-1 Style Intelligent Road Surface Disease Inspection Vehicle of Department of Computer Science of Nanjing University of Science and Technology. It aims to improve the system architecture and to study the automatic inspection algorithm of diseases in road surface images.The automatic inspection system of highway road surface is a complicated system. Firstly, the software and hardware environment, the system architecture, modules of the system and their functions are introduced. On this basis the system architecture is improved by providing a more effective and rational locating solution of road surface diseases. Then by analyzing the characteristics of the road surface images, we carry out some necessary preprocesses such as gray scale calibration and geometry calibration to these images.This paper is focused on crack disease detection. After reviewing and evaluating the traditional edge detection methods and pointing out their limitation in the detection of the road surface diseases, a new method based on hierarchical processing model is implemented. The hierarchical processing model is flexible and efficient and can meet the complicated situation of the road surface. The validity of the model is proved through experiments.