HAMMING DISTANCE BASED LOCAL BINARY PATTERN FOR BIOMEDICAL IMAGE RETRIEVAL

Image retrieval is a real world problem where we can extract images from the database on the basis of certain similarity. The application of image retrieval in the medical field can be very useful for the diagnosis of disease and further treatment planning. In this paper, we have modified the local binary pattern by calculating the transition between consecutive windows using hamming distance and then calculate the hamming code for building features. The feature set is used for retrieving images from the database by evaluating the similarity between features using distance metric. The experiments are conducted on real time database and the proposed algorithm shows significant improvement over existing local binary pattern based image retrieval algorithms