Prediction of wavelet transform coefficients using neural networks applied to lossless compression of multispectral images

We present a lossless compressor for multispectral images that combines two classical tools: wavelets and neural networks. Due to their huge dimensions, images are split into small blocks and the wavelet transform that maps integers to integers is applied to each block -and each band- to decorrelate...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Acevedo, D.G., Ruedin, A.M.C., Seijas, L.M.
Formato: CONF
Materias:
Lossless compression
Multispectral images
Neural networks
Prediction
Computation theory
Image compression
Integer programming
Wavelet transforms
Arithmetic coder
Acceso en línea:http://hdl.handle.net/20.500.12110/paper_0277786X_v6683_n_p_Acevedo
Aporte de:
Biblioteca Digital - Facultad de Ciencias Exactas y Naturales (UBA) de Universidad de Buenos Aires
id todo:paper_0277786X_v6683_n_p_Acevedo
record_format dspace
spelling todo:paper_0277786X_v6683_n_p_Acevedo2023-10-03T15:16:37Z Prediction of wavelet transform coefficients using neural networks applied to lossless compression of multispectral images Acevedo, D.G. Ruedin, A.M.C. Seijas, L.M. Lossless compression Multispectral images Neural networks Prediction Computation theory Image compression Integer programming Wavelet transforms Arithmetic coder Multispectral images Neural networks We present a lossless compressor for multispectral images that combines two classical tools: wavelets and neural networks. Due to their huge dimensions, images are split into small blocks and the wavelet transform that maps integers to integers is applied to each block -and each band- to decorrelate it. In order to increase even more the compression rates achieved by the wavelet transform, coefficients in the two finest scales are predicted by means of neural networks, which use causal information (ie, coefficients already coded) to get nonlinear estimates. In this work, we add coefficients from other spectral bands to compute the prediction, besides those coefficients belonging to the same band, which lie in a causal neighbourhood. The differences are then coded with a context based arithmetic coder. Several options regarding initialization, training and architecture of the neural networks are analyzed. Comparison results with other lossless compressors (with respect to the coding time and the bitrates achieved) are given. Fil:Acevedo, D.G. Universidad de Buenos Aires. Facultad de Ciencias Exactas y Naturales; Argentina. Fil:Ruedin, A.M.C. Universidad de Buenos Aires. Facultad de Ciencias Exactas y Naturales; Argentina. Fil:Seijas, L.M. Universidad de Buenos Aires. Facultad de Ciencias Exactas y Naturales; Argentina. CONF info:eu-repo/semantics/openAccess http://creativecommons.org/licenses/by/2.5/ar http://hdl.handle.net/20.500.12110/paper_0277786X_v6683_n_p_Acevedo
institution Universidad de Buenos Aires
institution_str I-28
repository_str R-134
collection Biblioteca Digital - Facultad de Ciencias Exactas y Naturales (UBA)
topic Lossless compression
Multispectral images
Neural networks
Prediction
Computation theory
Image compression
Integer programming
Wavelet transforms
Arithmetic coder
Multispectral images
Neural networks
spellingShingle Lossless compression
Multispectral images
Neural networks
Prediction
Computation theory
Image compression
Integer programming
Wavelet transforms
Arithmetic coder
Multispectral images
Neural networks
Acevedo, D.G.
Ruedin, A.M.C.
Seijas, L.M.
Prediction of wavelet transform coefficients using neural networks applied to lossless compression of multispectral images
topic_facet Lossless compression
Multispectral images
Neural networks
Prediction
Computation theory
Image compression
Integer programming
Wavelet transforms
Arithmetic coder
Multispectral images
Neural networks
description We present a lossless compressor for multispectral images that combines two classical tools: wavelets and neural networks. Due to their huge dimensions, images are split into small blocks and the wavelet transform that maps integers to integers is applied to each block -and each band- to decorrelate it. In order to increase even more the compression rates achieved by the wavelet transform, coefficients in the two finest scales are predicted by means of neural networks, which use causal information (ie, coefficients already coded) to get nonlinear estimates. In this work, we add coefficients from other spectral bands to compute the prediction, besides those coefficients belonging to the same band, which lie in a causal neighbourhood. The differences are then coded with a context based arithmetic coder. Several options regarding initialization, training and architecture of the neural networks are analyzed. Comparison results with other lossless compressors (with respect to the coding time and the bitrates achieved) are given.
format CONF
author Acevedo, D.G.
Ruedin, A.M.C.
Seijas, L.M.
author_facet Acevedo, D.G.
Ruedin, A.M.C.
Seijas, L.M.
author_sort Acevedo, D.G.
title Prediction of wavelet transform coefficients using neural networks applied to lossless compression of multispectral images
title_short Prediction of wavelet transform coefficients using neural networks applied to lossless compression of multispectral images
title_full Prediction of wavelet transform coefficients using neural networks applied to lossless compression of multispectral images
title_fullStr Prediction of wavelet transform coefficients using neural networks applied to lossless compression of multispectral images
title_full_unstemmed Prediction of wavelet transform coefficients using neural networks applied to lossless compression of multispectral images
title_sort prediction of wavelet transform coefficients using neural networks applied to lossless compression of multispectral images
url http://hdl.handle.net/20.500.12110/paper_0277786X_v6683_n_p_Acevedo
work_keys_str_mv AT acevedodg predictionofwavelettransformcoefficientsusingneuralnetworksappliedtolosslesscompressionofmultispectralimages
AT ruedinamc predictionofwavelettransformcoefficientsusingneuralnetworksappliedtolosslesscompressionofmultispectralimages
AT seijaslm predictionofwavelettransformcoefficientsusingneuralnetworksappliedtolosslesscompressionofmultispectralimages
_version_ 1807320972946046976

Ejemplares similares