Camera Calibration Algorithm Library Design And Experimental Verification

The research of the camera calibration already has been more than 20 years, and at present its research methods are mainly divided into two big categories: traditional method and self-calibration method. As far as traditional method, it takes the optimized algorithm and the perspective transformation matrix algorithm as representative that a target is put before the camera and the parameters of the camera are calculated according to the relationship between the reference points in three dimensional coordinates and the reference points in image coordinates. Traditional method is quite complex, but we can get a high precise result. Self-calibration doesn't need targets, however, it can calculate the parameters from the restriction conditions obtained from image sequence. It is Real-time and on-line, while it is less precise than traditional method.At present, the research of the camera calibration is to design and improve algorithms to improve the precision and the reliability of the result in order to get a more satisfying result in advanced work. In this paper the traditional method is the key point and three calibration algorithms are chosen, Perspective transformation matrix algorithm, Tsai algorithm and Zhang algorithm for that these algorithms are representative, characteristic and used widely. Further more, these algorithms are easy to understand.This paper presents following results through study on experiment of the algorithms above:1. Know camera calibration's effect on computer vision measurement.2. Know more about the algorithms and its original thermos by experiment and test.3. Design algorithm library of camera calibration and present the software package.4. Have an experimental flow for the computer vision measurement and accumulate some experiences.