A VLSI Architecture for Lifting-based 2-D Discrete Wavelet Transform

WANG Chao,CAO Peng,LI Jie,HUANG Weida
DOI: https://doi.org/10.3969/j.issn.1004-373X.2007.14.041
2007-01-01
Abstract:Discrete Wavelet Transform(DWT) is a computationally intensive process and the memory requirement for its implementation is also high.To make it is suitable for real-time image processing applications,it is essential to develop special-purpose architectures and custom VLSI chips for computation of DWT.In this paper,we propose a new lifting-based 2-D DWT architecture.The architecture is a hybrid of level-by-level and line-based architectures and has less hardware cost and memory requirement.It can be extended to multi-lifting steps and can be used for multi-level 2-D DWT.
What problem does this paper attempt to address?