A HYBRID COMPRESSION MODEL FOR CLUSTERS OF SIMILAR MEDICAL IMAGES

In this paper, a new compression scheme called hybrid compression model (HCM) is proposed for compressing clusters of similar images. The HCM exploits region growing to segment the median image that created by a cluster of similar images; and further, it uses centroid method to predict the values of original images data. The difference between the predict values and original data is stored for later use of progressive transmission. The experimental results obtained on various images show that our method provides significant improvement in compression efficiency while compared to the traditional centroid method.