Lossy To Lossless Medical Image Coding Using Joint Bit Scanning Method

Abstract

A new algorithm for progressive medical image coding is presented. On the 8-bit gray scale image, lifting based integer wavelet transform (IWT) are applied to get the three level multi-resolution Integer wavelet transformed image. Then, it is encoded using block based partitioning scheme to exploit the energy clustering in frequency and in space. Whenever a pixel is found significant, pixel value is completely transmitted using vertical bit scanning and then proceeds again with block based coding. Experiments are carried on MRI images to prove the effectiveness of the proposed algorithm. The results shows a significant improvement in terms of distortion measured as peak signal to noise ratio (PSNR) and Correlation Coefficient (CoC)  for a given bit rate compared to the existing state of the art embedded image coding methods.