You are here

A hardware implementation of discrete wavelet transform for compression of a natural image

TitleA hardware implementation of discrete wavelet transform for compression of a natural image
Publication TypeConference Proceedings
Year of Conference2017
AuthorsPadmavati, S., V. Meshram, and D. Jayadevappa
Conference Name2017 International Conference on Algorithms, Methodology, Models and Applications in Emerging Technologies, ICAMMAET 2017
Volume2017-January
Pagination1 - 5
Date Published2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN Number9781509033782 (ISBN)
KeywordsDepartment of Electronics and Communication Engineering, Scopus
Abstract

Digital images can be represented in different forms/representations by using Image Transforms, which are simple or complex mathematical operations on the image. These transforms finds application in compression, enhancement, pattern recognition, and feature extraction etc. Discrete Wavelet Transform (DWT) is a widely used wavelet transformation. DWT transforms image from space domain to frequency domain, where the wavelets are discretely sampled. DWT has various advantages over other image transforms such as image can be represented in multi resolution form, avoids the division of image into non-overlapping blocks, blocking artifacts are avoided to achieve a high compression ratio, better identification of data suitable for human perception, provides better localization in spatial and frequency domain. DWT implemented on software [Matlab] usually has lower throughput, which is due to high overall design cycle time. This can be improved through hardware implementation of DWT. In this paper an effective high speed DWT architecture has been designed using Distributed Arithmetic (DA) concept. The designed architecture has been simulated and synthesized by using Verilog HDL for the selected device ZYNQZC706 (7z045ffg900-2). Our implementation is validated on a natural image of dimension 512∗512. This implementation reduces the overall design cycle time to 0.670 nanoseconds. © 2017 IEEE.

DOI10.1109/ICAMMAET.2017.8186683
Short TitleInt. Conf. Algorithms, Methodol., Model. Appl. Emerg. Technol., ICAMMAET