节点文献
基于内存优化和位长搜索的小波分块编码算法
A wavelet block coding method using memory optimization and bit-length search
【摘要】 针对小波编码算法在编码过程需要占用大量内存的问题,提出一种基于内存优化和位长搜索的小波分块编码算法.该算法通过以下几项措施来减少编码过程的内存占用,同时提高编码的工作效率:采用较大的量化步长;提出位长定义,建立块最大绝对值的位长位图;以位长差的方式输出位长位图;分块独立编码,并采用一次性扫描输出.实验结果表明,该算法在保持高信噪比的同时,内存占用仅为8.27 Kbyte,有效地减少了编码过程的内存占用,为小波编码的硬件实现提供了一条有效的途径.
【Abstract】 Wavelet coders would need a lot of memories in coding process.A wavelet block coding method is proposed based on memory optimization and bit-length search,in which four measures were adopted to reduce the memory occupancy and improve the coding efficiency.These measured are:(1) to used the big quantization step length;(2) to define bit-length and establish block bit-length chart;(3)to output the bit-length chart is in terms of bit-length difference and block independent code and(4) to use scan output for one time merely.The experimental results show that the memory occupancy being 8.27 Kbyte only,the arithmetic can reduce memory occupancy in the process of coding and increase the coding speed effectively to keep the high quality.
【Key words】 block coding; memory optimization; bit-length chart; discrete wavelet transform(DWT); bit-length search;
- 【文献出处】 华中科技大学学报(自然科学版) ,Journal of Huazhong University of Science and Technology(Nature Science Edition) , 编辑部邮箱 ,2008年06期
- 【分类号】TN919.81
- 【被引频次】2
- 【下载频次】62