The Circle Detecting Based On Digital Image Processing

The quick and precise detection of circles has a broad prospect of applications in thefield of computer vision and automatic detection. The ways which are widely used atpresent are contact non-contact ones such as air micrometer. But they have many defectsincluding small scope, low precision, low resolution, inefficiency, and the inevitable errorcaused by human himself.In this thesis, an automatic circle detection system is developed based on digitalimage processing techniques, in which, CCD camera grabs the microscopic image ofto-be-measured end surface, video image grabber digitizes it and computer processes it bymeans of digital image processing and then gives the results.The present thesis focuses on studying and realizing the algorithm of circle detectionby means of programming the common algorithm of image processing. It also improvesthe Hough transformation based on the geometry feature of circles and applies it to themulti-circles detection in the images.