Detection Algorithm For Asphalt Pavement Crack Based On Image Processing

Pavement distress detection plays an important role in the highway management and maintenance. But the traditional method for the pavement distress detection is still used in our nation. This method has many disadvantages, which disatisfies with the requirement for the highway management and maintenance. Based on that the highway construction and transit is urgently in need of the automation management, the image detecting algorithms and system are discussed for the asphalt pavement crack in this thesis. The main work and accomplishment are as follows:(1) The gray of the crack target is close to that of the background region, so it is very hard to pick up the crack target. In order to solve this problem, the fuzzy set theory is introduced to crack detection for the asphalt pavement. A novel generalized fuzzy image enhancement operator with close-character and transplantable-character is proposed. Then the presented operator is applied to image enhancement algorithm, which can enhance the crack edge better.(2) In the practical applications, it is difficult to make a balance between processing effect and processing speed for the edge detection algorithm. In order to solve this problem, a novel edge detection algorithm is presented based on the mathematics morphology. The proposed algorithm employs the gradient operator to gain the interested regions in the image. Based on the morphology-gradient operator and OTSU segmentation operator, then the proposed algorithm constructs four structural elements to detect the edges in interested regions. This algorithm has not only a good ability to extract the image edge, but also a fast speed ability to process the image.(3) The extracted objects are marked and computed, and the marked noises are filtered. The interspaces of the crack are filled by region filling method. Then the gained cracks are classified as regular cracks and irregular cracks. At last, the length and width of the thinned regular crack are obtained by counting method, or the area of the irregular crack is obtained by nibbling method.(4) A novel crack detecting system for the asphalt pavement is proposed based on the image processing technology. In the programming environment of Visual C++ 6.0, the corresponding modules and image processing algorithms are implemented by dint of Mil 8.0 SDK. Finally, this system is applied to the asphalt pavement crack image detection experiment, and the experiment result is satisfactory. So the presented system has very high practicability and generality.