期刊名称：International Journal of Image Processing (IJIP)
电子版ISSN：1985-2304
出版年度：2011
卷号：5
期号：4
页码：371-381
出版社：Computer Science Journals
摘要：A common approach to the medical image compression algorithm begins by separating the region of interests from the background of the medical images and then lossless and lossy compression schemes are applied on the ROI part and background respectively. The compressed files (ROI and background) are now transmitted through different media of communications (local host, Intranet and Internet) between the server and clients. In this work, a medical image transfer coding scheme based on lossless Haar wavelet transforms method is proposed. At first, the proposed scheme is tested on Intranet (for both RoI and background) in order to compare its results with Internet tests. An adaptive quantization algorithm is used to apply on quasi lossless ROI wavelet coefficients while a uniform quantization is used to apply on lossy background wavelet coefficients. Finally, the retained quantization indices are entropy encoded with an optimal variable coding algorithm. The test results have indicated that the performance of the proposed MITC via Intranet is much better than via Internet in terms of transferring time, while the quality of the reconstructed medical image remains constant despite the medium of communication. For best adopted parameters, a compressed medical image file (760 KB "³ 19.38 KB) is transmitted through Internet (bandwidth= 1024 kbps) with transfer time = 0.156 s while the uncompressed file is sent with transfer time = 6.192 s.
关键词：Haar Wavelet; Region of Interest Coding; Adaptive Quantization
