CAJ | 학술논문

鉴于SURF算法能快速而稳定地提取图像特征,提出一种基于SURF特征进行快速标定相机焦距的方法.相机模型采用单参数模型（即只有焦距未知）,先使用SURF算法得到图像匹配点对,这些匹配点对用于计算基础矩阵,然后根据Kruppa方程推导出一个二次方程,进而求解出焦距.实验结果验证了此方法的有效性和正确性.
감우SURF산법능쾌속이은정지제취도상특정,제출일충기우SURF특정진행쾌속표정상궤초거적방법.상궤모형채용단삼수모형（즉지유초거미지）,선사용SURF산법득도도상필배점대,저사필배점대용우계산기출구진,연후근거Kruppa방정추도출일개이차방정,진이구해출초거.실험결과험증료차방법적유효성화정학성.
In view of the fact that SURF algorithm can rapidly and stably extract image characteristics, a method for camera focal length self - calibration based on SURF characteristics was proposed in this paper. The single parameter model was adopted as for camera model, that is, the focal length was not determined only. Firstly the matched- point sets were calculated by SURF algorithm to calculate fundamental matrix. Secondly a quadratic equation was educed according to the Kruppa equation, and then the focal length can be solved. The experimental results validated the validity and correctness of this method.