A approach to similarity retrieval in image database systems

• Main Authors: Huang, TE-WEI, 黃悌惟
• Other Authors: Liang, En-Hui
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/56093419900473385636

碩士 === 淡江大學 === 資訊管理學系 === 91 === The spatial relation between objects of images is one of important characteristics of images. Similarity retrieval is performed based on the spatial relation in many application of image data bases. For image indexing, many data structures derived from 2D strings have been proposed. 2D B-string is a symbolic representation of images, and similarity retrieval based on string matching between 2D B-string has been proposed. However, only if the same object in different images is at the same relative spatial location is considered and the difference between them is ignored. In this paper, three kinds of similarity between images and the ways of evaluating them are proposed to solve the problem stated above. The final similarity between images is obtained by combining those similarities, each of which with an adjustable weight. Finally, an experiment is performed to verify the effectiveness of the proposed approach and the analysis is also given.