Study On High Accuracy Locating System Based On Computer Vision

As one of the most important instrument for manufactory of microchip, the study of high accuracy locating system have become more and more popular.This paper has designed a self-Locating system based on computer vision, which is using Image processing method to realize the micro parts' high accuracy locating. The main work of the dissertation include:1. Studied on floating avoiding method and the memory conflict regulating method to realize the real-time image capture.2. Using four subpixel-locating methods to obtain the center of micro parts, which are template-matching method, core-shape method, centrobaric gray method, and edge-fitting method.3. Set up a coordinate transforming relation between image coordinate and dimensional coordinate.4. To improve the stability and precision of the system, median filter and filter based on morphologic method have been adopted.5. Finally, several experiments have been proposed to verify the availability of the system presented in this paper. Results show that the system can accurately detect the position of measured micro parts.