Nonlinear prediction with neural networks applied to 1-D DPCM image coding

Document Type

Article

Publication Date

1-1-1992

Abstract

Neural net architectures, with a hidden layer or functional links, have been utilized to generate prediction for one-dimensional (1-D) differential pulse code modulation (DPCM), applied to still image coding. In this approach, the predictor is designed by supervised training based on a typical sequence of pixel values; i.e., the values of the coefficients of the predictor are determined by training on examples. Nonlinear as well as linear correlations are exploited. Computer simulation experiments have been carried out to evaluate the resulting performance. At a transmission rate of 1 bit/pixel, for the images Lena and Baboon, the 1-D neural network DPCM provides a 4.17- and 3.74-dB improvement in peak SNR, respectively, over the standard linear DPCM system. © 1992.

Identifier

44049118097 (Scopus)

Publication Title

Journal of Visual Communication and Image Representation

External Full Text Location

https://doi.org/10.1016/1047-3203(92)90021-K

e-ISSN

10959076

ISSN

10473203

First Page

247

Last Page

254

Issue

3

Volume

3

This document is currently not available here.

Share

COinS