Medical Image Compression with Region of Interest and Segmentation

Abstract

The medical image compression, judgment is efficient only when compression techniques preserve all the important and relevant information needed. This can be with lossless compactness methods. Lossy compactness techniques can be well organized in terms of storage and transmission needs but there is no guarantee that they can provide the exact information necessary in medical image segment and diagnosis. The lossy compression, in which the information is usually considered in the coefficients of the domain space of original image. That is, the DWT based medical image compression, in which the wavelet coefficients maintain all the sequence needed for decompression of medical image. The objective of such a compactness methodology is that the maximization of the compression ratio. newlineThe segmentation of image is a key method for the image analysis. It is an effective technique for textured image region. The concept of texture gradient is implemented using the complex wavelet transform. Extensive researches efforts have been spent in the recent past on wavelet based image coding and the part of my work is aligned with these efforts. The algorithm produces an effective texture and intensity based segmentation for the application of medical image processing. The natural images have also taken to represent the result using resent coding algorithm into specific frequency domain like DCT, DWT etc. The results are obtained on the basis of coefficient s that is sufficient for reconstructing the image with good quality. newlineA comparative analysis of the results for the existing standard method on the basis of ROI by proposed algorithms using different methods like masking, mean-mean and max-min methods. That is enabling a combined ROI image. The Region of Interest which are priority during the segmentation in order to exhibit a higher quality than the rest. The split and merge algorithm are purposed to separate ROI and Non ROI of the image. newline newline