JPEG2000算术编码器的改进算法与实现

An Improved Algorithm for the Arithmetic Encoder in JPEG2000 and its Implementation

作为新一代静止图像压缩标准的JPEG2000标准拥有压缩比高,支持多分辨率等优点。JPEG2000的编码方式采用了嵌入式码块编码(EBCOT)方式,在编码过程中采用了MQ算术编码。文中分析了它对内容单一、信息量少图像编解码的不足,针对这些不足提出了一种对MQ算术编码器流程的改进算法。这种算法提高了JPEG2000对简单图像压缩的PSNR值,使解码后的图像更加清晰。

As a new still image compression standard,JPEG2000 has many advantages,such as higher compression rate and multi-resolution format.JPEG2000 coding is embedded block coding with optimized truncation（EBCOT）,and the process of code adopts the MQ arithmetic coding.This paper analyzes the deficiency of MQ arithmetic coder in coding the image with single content and less information,and proposes an improved algorithm of the MQ arithmetic coder to improve the PSNR of coding the simple images and the quality of the decoded images.