文献类型：期刊文献

中文题名：机器视觉技术中一种基于反对称矩阵及RANSAC算法的摄像机自标定方法

英文题名：An Approach of Camera Self-Calibration Based on Skew-symmetric Matrix and RANSAC alogrithm

作者：王赟[1]

第一作者：王赟

机构：[1]新乡学院机电工程学院

第一机构：新乡学院机电工程学院

年份：2015

卷号：0

期号：4

起止页码：92-94

中文期刊名：现代制造技术与装备

外文期刊名：Modern Manufacturing Technology and Equipment

语种：中文

中文关键词：摄像机自标定;基本矩阵;反对称矩阵

外文关键词：camera self-calibration,fundamental matrix,skew-symmetric matrix

摘要：介绍了一种摄像机自标定方法,该方法通过匹配的特征点建立标准矩阵后,利用反对称矩阵的性质,将标准矩阵表达式分解成6个约束方程,通过其约束关系得到摄像机内外参数。同时采用了RANSAC算法从检测到的特征点中排除奇异的特征点,对数据集进行筛选,以此提高匹配点的准确度和标定的精度。实验表明该方法能根据真实视频获得摄像机内外参数,能够较好的应用于机器视觉领域。
This paper describes a self-calibration method . After establishing fundamental matrix by using matched feature points , six constraints equations was founded from the fundamental matrix based on the character of the skew-symmetric matrix . . Then the intrinsic and ex-trinsic parameters can be determined through the relation of the set of constraints . Ransac method was adopted to exclude the singular points from detected feature points , therefore improve the accuracy of feature matching and camera calibration . Experimental results for real video showed that this method can effectively acquire the intrinsic and extrin-sic parameters , and it can be applied into computer vision field .