Journal on Communications ›› 2008, Vol. 29 ›› Issue (1): 132-140.doi: 1000-436X(2008)01-0132-09

• Academic communication • Previous Articles    

Coder selection based on context for wavelet image compression algorithm

Ying-hong ZHOU,Zheng-ming MA   

  1. Dept.of Electronic & Communications Engineering,Sun Yat-sen Univ.,Guangzhou 510275,China
  • Online:2008-01-25 Published:2017-07-01

Abstract:

A wavelet image coding algorithm based on context selection was proposed:compares the ideal average coding cost of arithmetic coding,fixed run-length coding and variable run-length coding in the case of independently-identically distribution; analyses the increase of actual coding cost due to the existence of error of estimated probability and time-varying probability ; builds a new context template; chooses arithmetic coding,fixed run-length coding or variable run-length coding based on the proposed context model.Experimental results show that speed-up results are achieved in the new algorithm,for example,at 0.25bit/pixel,the proposed scheme saves 20% above coding time compared with arithmetic coding scheme; the proposed algorithm improves 0~0.28dB over embedded block coding with optimized truncation(EBCOT) in terms of PSNR and especially suits for low bit rate compression application.

Key words: wavelet image coding, arithmetic coding, fixed run length coding, variable run length coding

No Suggested Reading articles found!