Abstrakt

Natural Language Querying for Content Based Image Retrieval System

Sreena P. H., David Solomon George

Content based image retrieval (CBIR) system is a database management system for retrieval of images based on the similarity of image content with the query image. The issue of semantic gap causes retrieval of irrelevant images from database. In the proposed CBIR system, fuzzy clustering of the Tamura texture features extracted from the database is used to overcome this problem. Natural language querying is implemented by use of fuzzy data space. The fuzzy membership values derived using fuzzy c-means clustering is used as similarity measure. The proposed technique is implemented in Matlab and its effectiveness is verified using the standard Brodatz texture database.