Skip navigation
Please use this identifier to cite or link to this item: https://libeldoc.bsuir.by/handle/123456789/41685
Title: Thresholding Neural Network Image Enhancement Based on 2-D Non-separable Quaternionic Filter Bank
Authors: Avramov, V. V.
Rybenkov, E. V.
Petrovsky, N. A.
Аврамов, В. В.
Рыбенков, Е. В.
Петровский, Н. А.
Keywords: авторефераты диссертаций
image enhancement
thresholding neural network
Issue Date: 2019
Publisher: Springer, Германия
Citation: Avramov, V. V. Thresholding Neural Network Image Enhancement Based on 2-D Non-separable Quaternionic Filter Bank / V. V. Avramov, E. V. Rybenkov, N. A. Petrovsky // Pattern Recognition and Information Processing : 14th International conference, Minsk, 21–23 may 2019 / Springer. – Minsk, 2019. – P. 147–161. – DOI: https://doi.org/10.1007/978-3-030-35430-5_13.
Abstract: The thresholding neural network with a 2-D non-separable paraunitary filter bank based on quaternion multipliers (2-D NSQ-PUFB) for image enhancement is proposed. Due to the high characteristics of the multi-bands 2-D NSQ-PUFB (structure “64in-64out”, CG2D=17,15dB , prototype filter bank ( 8×24 ) Q-PUFB), which forms the basis of the TNN, the results of noise editing in comparison with the approaches based on the two-channel wavelet transform in terms of PSNR are 1dB – 1.5dB higher.
URI: https://libeldoc.bsuir.by/handle/123456789/41685
Appears in Collections:Публикации в зарубежных изданиях

Files in This Item:
File Description SizeFormat 
Avramov_Thresholding.pdf21,91 kBAdobe PDFView/Open
Show full item record


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