文章基本信息

标题：A Robust Estimation Algorithm of Epipolar Geometry through the Hough Transform
本地全文：下载
作者：Shan Haitao ; Hao Xiangyang ; Wang Hui 等
期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2008
卷号：XXXVII Part B2
页码：985-990
出版社：Copernicus Publications
摘要：Fundamental matrix is an algebraic expression of the epipolar geometry, and plays a very important role in the computer vision. It is all along an important research topic that how to raise the precision and robustness of the fundamental matrix estimation in the field of computer vision. On the basis of this point, this paper firstly analyzes the character and function of the fundamental matrix, and proves that if there is more matching points than three in a single two-dimension line, the extra points will be regarded as redundant points in theory. Those points will not be used as matching points. Then in order to eliminate those collinear redundant points at the same time of establishing the corresponding relationship among those matching points, especially on images which there are many line features, the author introduces the Hough transform algorithm. Finally, the experiments show the average epipolar distance and residual errors of the fundamental matrix which is estimated with the improved algorithm in which the redundant points are eliminated is less than that of using the previous algorithm .The accuracy and robustness of estimation are improved evidently
关键词：Computer vision; Fundamental Matrix; Epipolar Geometry; Collinear Point Redundancy; the Hough Transform