摘要
在传统的图像小波变换中,首先需要对每行进行小波变换,等所有的行变换完成后再进行列小波变换,然后再进行嵌入式量化和编码,这就需要足够的存储容量保存这些变换结果,而且运算速度也慢。该文提出了一种基于小波变换的快速编码方法。在行变换中,仍然使用传统的提升结构,在列变换中,使用基于因果系统的低存储容量的提升结构。一旦有小波系数输出,就进行相应的量化、编码。该方法不仅压缩效果较好,而且在运算时间上得到大大提高。
In the conventional wavelet transform,all the lines are filte re d before column wavelet transform starts and then embedded quantization and enc oding are re quired.The fast compression method using line-based wavelet trans form is presented.In row transform,we use the conventional lifting scheme ,a nd in column we use low-memory lifting structure with causual system.In ord er to keep the memory needs low,wavelet coefficients are com pressed soon after they have generated.The experimental results demonstrate the coding performanc e is effective,and it offers a significant advantage in terms of speed and me mory needs.
出处
《计算机工程与应用》
CSCD
北大核心
2003年第33期45-46,158,共3页
Computer Engineering and Applications
基金
北京市教育委员会科技发展计划项目(编号:01KJ034)
关键词
第二代小波变换
提升方案
图像编码
second generation wavelet transform,lifting scheme ,ima ge encoding