Grid approach for X-ray image classification

The process of medical image classification is still carried out manually using the knowledge of the physician or radiologist, which leads to inaccurate and slow process of object identification. Thus, we need an automatic system that can classify medical images, accurately and faster from query images into one of the pre-defined classes. In this research, we are dealing with the classification of medical image to the image classes that are defined in the database. We focus on using the shape of X-ray image to carry out the classification process and to use the Euclidean distance and Jeffrey Divergence techniques to measure image similarity. In this paper, we use a grid approach to simplify the shape of X-ray images to obtain a better recognition rate. Our experiment shows that this approach gives a higher recognition rate.