Content based Mammogram Image Retrieval using Particle Swarm Optimization and Hybrid Classifier Arulmurugaselvi N.*, Dr. Jeyabharath R.**, Dr. Veena P.*** *Lecturer, Electronics and Communication Engineering Department, Government Polytechnic College, Coimbatore, India. arulphdresearch@gmail.com **Professor & Head, Electrical and Electronics Engineering Department, K.S.R Institute for Engineering and Technology, Tiruchengode, India. jeya_psg@rediffmail.com ***Professor, Electrical and Electronics Engineering Department, K.S.R Institute for Engineering and Technology, Tiruchengode, India.veena_gce@yahoo.co.in Online published on 23 March, 2017. Abstract In this paper present a content-based image retrieval system is designed to retrieve the mammogram images from the large medical image database. The proposed CBIR system consists of three steps. The first step is feature extraction based on shape, marginal and texture features, and the second is a selection of possible features using PSO algorithm. Finally, the third step is the retrieval of mammogram image using the hybrid classifier, a combination of artificial neural network (ANN) with Multilayer Feed-Forward Back propagation (MLFFB). Finally retrieval the mammogram images using MLFFB-ANN based on possible selection features. Furthermore, the results also show that the proposed learning approach of MLFFB-ANN can improve their retrieval performance compared to SVM and Naïve Bayes algorithm. Top Keywords Mammogram, PSO Algorithm, content based image retrieval, Multilayer Feed-Forward Back propagation, and artificial neural network. Top |