Application of an improved version of JPEG2000 based on mathematical morphology to medical images

The JPEG2000 is the latest presently image compression standard presented by international standards organization, which enables compressed image at very low bitrates. At a very low bitrate, the reconstructed image provides bad visual quality. At this rate, a variety of artifacts can be distinguished, and the reconstructed image provides bad visual quality. In order to enhance the visual quality of reconstructed image, we propose a compression method based on morphological filtering. This method decomposes the image data into low and high frequency subimages. These subimages are compressed by JPEG2000 encoder by assigning different bitrates to each subimage. Through decoding by the morphological filter, proposed method, applied to still medical images provides better visual quality over direct compression of the image data.