Please use this identifier to cite or link to this item:
Title: Increasing the Speed of Wavelet Image Processing with Decimation Using the Winograd Method
Authors: Lyakhov, P. A.
Ляхов, П. А.
Nagornov, N. N.
Нагорнов, Н. Н.
Semyonova, N. F.
Семенова, Н. Ф.
Abdulsalyamova, A. S.
Абдулсалямова, А. Ш.
Keywords: Computational complexity;Winograd method;Decimation;Wavelet transform;Digital filtering;High-performance computing
Issue Date: 2023
Citation: Lyakhov, P.A., Nagornov, N.N., Semyenova, N.F., Abdulsalyamova, A.S. Increasing the Speed of Wavelet Image Processing with Decimation Using the Winograd Method // Proceedings of the Seminar on Signal Processing, SoSP 2023. - 2023. - pp. 79-82. - DOI: 10.1109/IEEECONF60473.2023.10366139
Series/Report no.: Proceedings of the Seminar on Signal Processing, SoSP 2023
Abstract: Wavelet processing is actively used for image de noising, compression, and fusion. The high growth rate of quantitative and qualitative characteristics of digital images, lead to the need to increase the speed of wavelet image processing methods and their efficient implementation on modern hardware devices. This paper proposes the Winograd method (WM) to speed up the wavelet image processing methods due to group pixel processing. The scheme of wavelet filtering of images using WM with decimation has been developed. The proposed approach reduced the asymptotic computational complexity of wavelet transform up to 53 % compared to the pixel-by-pixel processing. Estimating the time spent on a hardware device based on a unit-gate model showed that WM reduces device delay up to 67% compared to the direct method. The proposed approach implementation of wavelet image processing on a field-programmable gate arrays and an application-specific integrated circuits is a promising direction for further research.
Appears in Collections:Статьи, проиндексированные в SCOPUS, WOS

Files in This Item:
File SizeFormat 
scopusresults 2956 .pdf
  Restricted Access
132.35 kBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.