Title :
Well-distributed SIFT features
Author :
Song, Rachel ; Szymanski, Janusz
Author_Institution :
Dept. of Electron., Univ. of York, York
Abstract :
A method to enhance the recognition of spatially distributed features, based on the scale invariant feature transform (SIFT), is reported. The key idea is to modify the way in which the selection of a set of contender interest points from each input image is carried out, using a non-maximal suppression approach in the different scale spaces.
Keywords :
feature extraction; image matching; image registration; SIFT feature; image matching; image registration; nonmaximal suppression approach; scale invariant feature transform; spatially distributed feature recognition;
Journal_Title :
Electronics Letters
DOI :
10.1049/el.2009.2954
