Huang, C.-T.C.-T.HuangTseng, P.-C.P.-C.TsengLIANG-GEE CHEN2018-09-102018-09-10200513875485https://www.scopus.com/inward/record.uri?eid=2-s2.0-17444386385&doi=10.1007%2fs11265-005-4960-4&partnerID=40&md5=0944b879f2fe4644f78da4043902676dhttp://scholars.lib.ntu.edu.tw/handle/123456789/313891In this paper, a VLSI architecture for lifting-based shape-adaptive discrete wavelet transform (SA-DWT) with odd-symmetric filters is proposed. The proposed architecture is comprised of a stage-based boundary extension strategy and the shape-adaptive boundary handling units. The former could reduce the complexity of multiplexers that are introduced to solve the shape-adaptive boundary extension. The latter consists of two multiplexers and can solve the shape-adaptive boundary extension locally without any additional register. Two case studies are presented, including the JPEG 2000 default (9, 7) filter and MPEG-4 default (9, 3) filter. According to comparison results with previous architectures, the efficiency of the proposed architectures is proven. © 2005 Springer Science + Business Media, Inc.Boundary extension; Discrete wavelet transform; Shape-adaptive; VLSI architectureBinary codes; High pass filters; Image analysis; Image quality; Multimedia systems; Signal processing; Signal to noise ratio; Textures; Wavelet transforms; Boundary extension; Discrete wavelet transform; Shape-adaptive; VLSI architecture; VLSI circuitsVLSI architecture for fifting-based shape-adaptive discrete wavelet transform with odd-symmetric filtersjournal article10.1007/s11265-005-4960-42-s2.0-17444386385