Image Searching Using Heuristic Method for Image Retrieval System

Abstract

Heuristic method is introduced as one of solutions for efficient searching process in . Content Based Image Retrieval System. This method deals with structural database which is built in indexing phase. The structure is graphformed and also known as landscape. Each node represents an image and each edge is labeled with similarity value as nearness prediction between connected nodes. The similarity value is calculated using combination of three image features, those are color, shape, and texture. In retrieving phase, the heuristic searching moves along these nodes with Breadth-First Search mechanism under certain constraints until meeting stop criterion. Each visited node that fulfills the requirements will be retrieved as searching result. The result shows that implementation of this method might fasten retrieval process up-to nine times faster with keeps on maintaining the relevancy of retrieval result.