Optimize ASIC design by statistics analysis and fixed parameter for special-imagery compression
Xinfeng Xu, Yong Hei · 2010 3rd International Conference on Biomedical Engineering and Informatics · 2010
Medical imagery needs the function of lossless compression to minimize image size to facilitate transmitting and storage. With high-speed and high-efficiency characteristics, FELICS algorithm has potential capability to do this job well. The research in this paper is not only based on FELICS algorithm, but also explores a practical method of fixing one of FELICS's dynamic parameters as constant to optimize its ASIC implementation for special medical imagery. That is, toggling this key parameter of Rice coding in FELICS to a fixed value is able to save chip's hardware resources, like die area, power consumption and computational workload. More important in this process is applying statistics analysis on real images to choose the biggest-probability value for it. At last, the fixing-k method has been validated by excellent results.