基于分解反折结构的低内存低复杂度离散小波变换

Low-memory and reduced-complexity discrete wavelet transform with decomposed flipping structure

通过将反折结构分解为奇、偶时间索引对应的操作,在不增加计算复杂度的前提下提出了一种内存需求更低且处理器负载均衡的即时DWT实现—分解反折结构DFS。以图像/视频压缩中常用的CDF 9/7小波滤波器组为例,DFS与提升结构LS及FS具有相同的计算复杂度,但是内存需求(单层变换)从6个存储单元下降为5个。实验结果表明,基于DFS的DWT分解相对于常规LS实现及实时LS实现分别加速了44%和14%。

By decomposing the Flipping Structure （FS） into even phase and odd phase,this paper proposes an on-the-fly implementation of DWT called Decomposed Flipping Structure （DFS）,which is characterized by low-memory budget,low computational complexity and balanced workload.For the CDF 9/7 wavelet filterbank popular in wavelet-based image/video coding schemes,the DFS has the same computational complexity as the FS,while the memory requirement reduces from six memory cells to five ones.The experimental results show that the proposed implementation has a speedup of about 44 ％ and 14％,respectively,compared with the traditional lifting scheme and on-the-fly LS.