A MEDICAL IMAGE RETRIEVAL SYSTEM BASED ON SEMANTIC ANNOTATIONS

摘要

This paper presents the design and implementation of asemantic Content Based Image Retrieval Systems (CBIR)developed in Matlab from scratch by choosing acombination of texture, color and shape as low levelfeatures to represent the images, and by using a multilabelingclassifier to associate these low level features to asemantic label. We used the Bayes Point machineclassifier to classify the images. The classification resultsare further enhanced by using an explicit relevancefeedback algorithm. The system is tested on a set ofmedical images combined with other types of images andthe results are presented.