Semantic Content Based Medical Image Retrieval Using Invariant Contourlet Features with Relevance Feedback Techniques

Feature extraction is a special form of dimensionality reduction. When the input data to an algorithm is too large to be processed and it is suspected to be redundant (much data, but not much information) then the input data will be transformed into a reduced presentation set of features (also named features vector). Transforming the input data into the set of features is called features extraction. If the features extracted are carefully chosen, it is expected that the features set will extract the relevant information from the input data in order to perform the desired task using this reduced representation instead of the full size input. This work presents a novel method for content-based image retrieval based on interest points with transformation techniques as Non-Sub sampled Contourlet Transform (NSCT). Interest points are detected from the scale and rotation normalized sub- image. With robustness to the image’s rotation, scale and translation, local features of every sub-band images are extracted to describe the image and make the similarity measure. Further, relevance feedback technique is used to bridge the gap between low levels features and high level concepts. The proposed method is tested on a large medical image database which shows a significant improvement in precision and average retrieval rate (ARR) with relevance feedback.