Research Of Rubber Bolt Surface Inspection Based On Visual Detection System

Visual detection system is a new technology that developed forty years ago, and it has been widely used in areas of our lives and industries, especially for industry inspection. Compared to the traditional inspection methods which are carried out by human being, visual detection system are more effective and accurate.This research adopted visual detection system to inspect surface quality of rubber bolt, and six chapters are included as follows.In chapter 1, the background and significance of the research were advanced. Then, the development and applications of visual detection system both at home and abroad were analyzed. The structure of the whole visual detection system, especially the hardware part was presented. At the end of this chapter, the task of the dissertation and the rubber bolt were shown.In chapter 2, the theory of CCD and CCD camera was introduced. And then the VFW (Video for Windows) and DirectShow were compared on capturing image based on CCD camera and why the VFW was chosen was clearly explained. At the same time, the method of capturing and saving the images with VFW was shown. The principle of the entire surface detection was also introduced.In chapter 3, the images captured were processed according to their characteristics. The methods of image processing included gray transformation, threshold segmentation, logic calculation, spatial filter and so on. They were all analyzed in great detail. Also some algorithms of these methods were shown. The experiments were implemented to check these methods.In chapter 4, the image rotation methods based on angle-information and least-distance were analyzed. Their principles and realizations were presented and compared in detail. They were also compared based on the experiments.In chapter 5, the surface quality of the rubber bolt was checked out by the image matching ratio and maximal disfigurement ratio. The image matching ratio was calculated by the whole rubber bolt matching ratio and ratio of the width and height. Then the maximal disfigurement of the surface was inspected by regional marking, and the ratio calculation of the surface was analyzed to check out the quality of the rubber bolt. At the end, the experiments integrating these methods were maken to check out the rubber bolt.In the last chapter, a conclusion of this dissertation is made and the prospects and limitations of the visual detection system are also indicated.