摘要
文章研究了一种适合在高速DSP上运行的快速零树编码算法。通过分析传统EZW的数据压缩流程以及DSP的处理机制,提出了一种新的压缩编码算法。本算法从大量图像小波变换系数的统计特性中提取先验知识,用以指导压缩编码,使之在DSP上运行的速度得到了极大的提高,并已在某高输入数据率的图像压缩项目上使用,效果良好。
Based on the analysis of data encoding procession of the classical EZW and the operation methods of DSP, a new fast algorithm of zerotree wavelets coding which adapted to high-speed DSP （ Digital Signal Processor） is considered in this paper. In the algorithm, apriori knowledge is extracted from statistical properties of image wavelet transform coefficients in order to guidance the compression coding. The new fast image compression algorithm is improved to be operated on the DSP in much faster speed. This new algorithm has been applied to an actual high-speed real-time compression project successfully and the result is very satisfied.
出处
《激光与红外》
CAS
CSCD
北大核心
2007年第5期485-489,共5页
Laser & Infrared
基金
国家863计划项目(No.2002AA813032)
关键词
图像压缩
实时处理
小波变换
DSP
EZW算法
image compression
real-time procession
wavelet transform
DSP
EZW algorithm
