作者: Ahmed H. Tewfik , Enhui Yang , Bin Zhu
DOI: 10.5281/ZENODO.36231
关键词:
摘要: Arithmetic coding is usually implemented in fixed precision. Such an implementation cannot efficiently code sources, such as image algorithms, that locally produce a small fraction of large alphabet symbols. In this paper, we propose novel approach to overcome inefficiency. The proposed algorithm uses dual symbol sets: primary set contains the symbols have occurred recent past and secondary all other Both sets are dynamically adapted local statistics. We summarize analysis describe results obtained by applying it images.