Entropy Driven Bit Coding For Image Compression In Medical Application

Abstract

This paper present a new bit redundancy coding for image compression in binary bit planar coding. In the compression model, images are coded into binary level to stream over a communication medium or to store the processed data in a remote location. In this processing, the representative coefficients are coded in binary level and to minimize the resource overhead these bits are compressed using binary compression logic. Among different coding logic, Huffman coding is the standard coding approach, standardized by JPEG and JPEG-2K image compression committee. The coding schemes computes a bit pattern occurrence probability and derive a allocating code word for a pattern to be compressed. The advantage of utilizing redundancy coding result in higher compression. However, the over-coding issue for lower pattern probability gives a inverse compression affect in image compression. In this paper, this issue is addressed an a new hybrid image coding approach developing a mixed model approach of variable entropy coding and fixed pattern allotment is proposed. The proposed hybrid approach termed “Selective Hybrid Coding” is used for the compression of medical samples and compared for performance evaluation to the conventional image compression model.