Extracting Region of Interest for lossless Medical Image Compression Based on DWT Transform and RLE Code

Abstract

Recently, medical images gathered significant amount of great interest due to its role in different medical applications. The main aim of this work is to produce minimum possible compressed size and the transmission time. In this paper, to reduce transmitted or stored image size, Region of Interest (ROI) is extracted to be compressed by hybrid techniques of lossless image compression. Discrete Wavelet Transform (DWT), Zigzag searching Algorithm and Run Length Encoding (RLE) techniques are combined to provide best compression performance. For performance assessment of the proposed model, set of measures are adopted in this work such as compression and decompression time compressed image size, Peak Signal to Noise Ratio (PSNR). Experiments of this work were conducted on MiniMammographic Database (MIAS) images. Yielded results provide considerable reduction of the compressed image by 39% with better performance regarding Compression Ratio (CR) and PSNR.