DocumentCode :
1992720
Title :
Knapsack formulation of image matching
Author :
Daskalakis, T.N. ; Daskalakis, C.N.
Author_Institution :
Dept. of Electr. Eng., Victoria Univ. of Manchester, UK
fYear :
1991
fDate :
14-17 Apr 1991
Firstpage :
2365
Abstract :
A technique is described to solve one of the problems encountered in an image analysis system, namely, region matching. The approach adopted is that of a medium level process. The information gathered during a segmentation stage is used to transform the original images into a collection of two-valued ones and a collection of gray-level histograms. Mathematical morphology is then used to define a measure of similarity on which the process is based. Correspondence are established between the segments according to a number of principles which are presented. A well behaved and efficient algorithm is proposed for the solution of the problem
Keywords :
picture processing; algorithm; gray-level histograms; image analysis system; image matching; image segmentation; knapsack formulation; mathematical morphology; medium level process; region matching; Biomedical measurements; Humans; Image analysis; Image matching; Image segmentation; Information analysis; Information systems; Laboratories; Morphology; Multimedia systems;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1991. ICASSP-91., 1991 International Conference on
Conference_Location :
Toronto, Ont.
ISSN :
1520-6149
Print_ISBN :
0-7803-0003-3
Type :
conf
DOI :
10.1109/ICASSP.1991.150795
Filename :
150795
Link To Document :
بازگشت