Title :
Image compression by a variable block size BTC technique using Hopfield neural networks
Author :
Qiu, Guoping ; Varley, M.R. ; Terrell, T.J.
Author_Institution :
Dept. of Comput. & Electron., Univ. of Central Lancashire, Preston, UK
Abstract :
The authors have developed a Hopfield (1982) neural network block truncation coding (BTC) technique, which has been shown, in the mean squared error sense, to improve coding performance over established BTC techniques. A new development in using a Hopfield neural network BTC scheme for image compression is described. Simulation results are presented which show that the new technique provides improved coding efficiency and maintains good visual quality in the reconstructed images
Keywords :
data compression; image coding; neural nets; Hopfield neural networks; block truncation coding; coding efficiency; coding performance; image compression; mean squared error; reconstructed images; simulation results; variable block size; visual quality;
Conference_Titel :
Neural Networks for Image Processing Applications, IEE Colloquium on
Conference_Location :
London
