A Content-Based Retrieval System for Endoscopic Images

Abstract

Based on the research of low-level visual image features including color clustering, color texture and shape, a new image retrieving method using multi-feature fusion and relevance feedback to retrieval images is proposed. By setting up a prototype system to evaluate the performance of the proposed method, the results illustrate that the method can retrieve endoscopic images more effectively, accurately and quickly than the one based on single feature because of its flexible ability to combine features and interactive relevance feedback.