Automatic Detection Of Road Surface Disease Research And System Implementation

With the development of high way, the research and developing of high way road surface detection system is very exigent. The system is very complex, of which detecting road's surface, road distress position and structure system design are three important technologies.By analyzing the type and cause of road distress, we known that road distress is the initial stage of road disease and pay more attention on road distress detection. The general flow of road distress detection based on image is: image sampling â†' calibrate and restore image â†' binarize image â†'reduce noise â†' classify and get distress. According to the imaging character of CCD, a new method to calibrate the distortion image is brought forward. The rule and method of getting threshold value are summarized. And a new method is implemented to binaize the calibrated image. Then the mathematical morphology method is used to reduce noise of image. At last, according to the line character we designed the classifier to obtain the object image including linear distress and nonlinear ones. The grey of road image is different from the distress, so we can obtain the distress.In the paper, a network synchrocyclotron based on client/server is implemented. The location of road distress can be obtained by using time mapping and distance mapping. At last the system structure is described. The whole system is comprised of two parts. One is the depart in the vehicle, the other is off-line data process system. Some basic rules in designing are put forward at the end of paper.