Title :
Using constraint inequality on estimated correlation for rapid image search
Author :
Kaneko, Shun´ Ichi ; Mae, Takahiro ; Miyamoto, Atsushi
Author_Institution :
Graduate Sch. of Eng., Hokkaido Univ., Sapporo, Japan
Abstract :
A rapid algorithm of searching for similar images to an object image in a set of registered images is proposed. By providing correlations of all the pairs of registered images in advance, we can derive a constraint inequality for similarity of the object. Using the upper bounds defined by the inequality, three conditions of neglect can be derived, which enable one to quicken the computation of searching. An efficiency rate of eight times in average could be obtained for searching 204 images in the set of 1200 registered images.
Keywords :
constraint theory; correlation methods; image registration; image retrieval; visual databases; binary images; constraint inequality; correlations; image database; image registration; image retrieval; lower bounds; registered images; similar image search; upper bound; Content based retrieval; Hardware; Image databases; Image registration; Image retrieval; Information retrieval; Laboratories; Pattern recognition; Production engineering; Robots;
Conference_Titel :
Pattern Recognition, 2002. Proceedings. 16th International Conference on
Print_ISBN :
0-7695-1695-X
DOI :
10.1109/ICPR.2002.1044619
