Identification of Similar Gastrointestinal Images through Content Based Image Retrieval System based on Analytical Hierarchical Process

Abstract

Content based image retrieval (CBIR) is a technique for automatically retrieved images from the large image repository on the basis of image features. Manual operation which is assigned to individual image features as well as equal weight distribution among themselves creates a matter of concern towards getting better performance. Such weight assignment task has been automated in this paper with the help of analytical hierarchical process (AHP). The objective of the study focuses on the proper weightage assignment to features as per the nature of image. The model has been implemented and tested for the gastrointestinal images containing eight different classes taken from $K_{vasir}$ medical database. The accuracy of the proposed CBIR system is high in terms of Precision and Recall over manual assignment. Hence, this retrieval system for searching similar gastrointestinal images can assist the physicians in identifying accurate gastrointestinal disease. This system can also become very useful for classifying hyper spectral and multi spectral images.