Pretreatment Research Of Camera Calibration

Camera calibration is a subject with both theoretic significations and practical importance which can be widely applied to country economy, science research and national defence construction. It's a focus in computer version field. The pretreatment research of camera calibration is discussed in the paper which mainly contains two parts: research of corner detection algorithms and regulation theory for distorted image, and then design a camera calibration experiment with OpenCV.Two parts are mainly developed in this paper:(1)Based on many experiments and data analysis of traditional Harris algorithm, we find a method to adjust the factors affecting greatly, and then put forward a adaptive Harris corner detection algorithm. Treat the distorted triangle chessboard as target image to mark the corners. Results says that the advanced Harris algorithm avoids the corner gathering and corner losing effectively ,detect the right corners correctly and rapidly, especially for the smooth curve lines whose curvature are within certain ranges. The great results of adaptive Harris algorithm also can be shown on the edge detection later.(2)The adaptive Harris algorithm can be applied to regulation for image distortion , camera calibration and rotary 3-D reconstruction. For the first one, an expression model of image distortion is created based on the coordinate of triangle barycenter; for the second application, a camera calibration programming experiment is executed combined Visual C++6.0 with OpenCV which includes corner detection, image regulation and interior and external factors, to prepare for more experiment; for the last, the adaptive corner detection has applied to the 3-D reconstruction.