Title :
Matching patterns of line segments by eigenvector decomposition
Author :
Chan, H.B. ; Hung, Y.S.
Author_Institution :
Dept. of Electr. & Electron. Eng., Hong Kong Univ., China
Abstract :
This paper presents an algorithm for matching line segments in two images which are related by an affine transformation. Images are represented as patterns Of line segments. Areas defined between line segments are used to describe a line pattern in the form of a proximity matrix. Matches are determined by comparing the feature vectors obtained from eigenvalue decomposition of the proximity matrices. Reliable matches of line segments are obtained in both synthetic and real images
Keywords :
computer vision; eigenvalues and eigenfunctions; feature extraction; image matching; image representation; image segmentation; matrix decomposition; affine transformation; computer vision; eigenvalue decomposition; eigenvector decomposition; feature vectors; image matching; image representation; line segment patterns; proximity matrix; Computer vision; Eigenvalues and eigenfunctions; Equations; Image edge detection; Image segmentation; Matrix decomposition; Pattern matching; Testing; World Wide Web;
Conference_Titel :
Image Analysis and Interpretation, 2002. Proceedings. Fifth IEEE Southwest Symposium on
Conference_Location :
Sante Fe, NM
Print_ISBN :
0-7695-1537-1
DOI :
10.1109/IAI.2002.999934
