CBIR-MD/BGP: CBIR-MD System based on Bipartite Graph Partitioning

Content based image retrieval system for medical images is a method of retrieving medical images based on similarity of their visual contents. An efficient CBIR-MD system can help the doctors in retrieving similar medical image from the dataset to diagnose the disease efficiently. In this paper, a system is proposed in which query image is divided into equal size sub-blocks. The feature extraction of each sub-block is carried out using Haar wavelet and Fourier descriptor. A matching scheme based on Most Similar Highest Priority (MSHP) principle and the adjacency matrix of bipartite graph partitioning (BGP) formed using sub-blocks of query and target image, is provided for matching the image. The performance of proposed system is investigated in terms of precision-recall.