Title :
Research on Image Automatic Registration Based on Heterogeneous Features
Author :
Lu, Bei ; Wang, Xiaohua ; Yang, Bing
Author_Institution :
Inst. of Comput. Applic. Technol., Hangzhou Dianzi Univ., Hangzhou, China
fDate :
March 31 2009-April 2 2009
Abstract :
The image registration based on the features of the image is a major research direction.The points feature is used in traditional registration methods,but the lines and other basic features are ignored. So,an image auto registration method based on natural heterogeneous features is given out. First, the feature points are detected using Harris feature detector and the feature line segments are detected using Hough transfer, then the one to one matching points and lines could be obtained through rough match, then the points feature and the parameters of lines is used to match refinely, finally, the image registration could be achieved through the homography matrix calculated by matching corner points and the parameters of line segments.
Keywords :
Hough transforms; feature extraction; image matching; image registration; matrix algebra; Harris feature detector; Hough transfer; corner point; feature line segment; homography matrix; image automatic registration; natural heterogeneous feature; rough match; Computer applications; Computer science; Computer vision; Detectors; Geometry; Image processing; Image registration; Image segmentation; Object detection; Remote sensing;
Conference_Titel :
Computer Science and Information Engineering, 2009 WRI World Congress on
Conference_Location :
Los Angeles, CA
Print_ISBN :
978-0-7695-3507-4
DOI :
10.1109/CSIE.2009.265