Content Based Image Retrieval Using Fuzzy Relaxation and Rotational Invariance for Medical Databases

Newer generations of diagnostic machines are based on digital technologies for data acquisition and consequently with the emergence of digital archiving systems for preservation of diagnosis is rapidly increasing. The goals of Content-Based Image Retrieval (CBIR) systems is to process on collections of images, extract features and based on visual queries, extract relevant image from a repository. The fuzzy relaxation pattern matching technique using rotational invariance has been developed in the framework of fuzzy set and possibility theory. Our method takes into account the uncertainty arising in the calculated values which have to be compared for content based image retrieval. This paper looks into the basic principles and extends to the fuzzy relaxation technique for a given tolerance. Our method was extensively tested in a medical database to identify the class in the classification problem.