A novel approach for content based image retrieval in context of combination S C techniques

This paper presents content based image retrieval (CBIR) system using immense power of combination of soft computing techniques such as artificial neural network, fuzzy logic and support vector machine. Traditional content based image retrieval system are retrieved images using low level visual features, hence, suffer from semantic gap. There is need to reduce semantic gap and improve accuracy of CBIR even though size of image databases increase rapidly, because accuracy and efficiency of content-based image retrieval system is extremely important in any real world application. Proposed innovative framework for content-based image retrieval based on combination of three soft computing techniques enormously improves accuracy of image retrieval. A user gives input to the system in the form of specified query image and system return set of relevant images. Proposed system employs relevance feedback based on SVM that intelligently classify images relevant or irreverent to given query image. Performance of proposed content based image retrieval system pragmatically evaluated in term of precision, recall and accuracy.