Neural net based digital halftoning of images
- 6 January 2003
- proceedings article
- Published by Institute of Electrical and Electronics Engineers (IEEE)
Abstract
Various novel techniques for digital image halftoning are presented, performing nonstandard quantization subject to a fidelity criterion. Hopfield-type networks can be used for this task, minimizing a frequency-weighted mean squared error between the input (continuous-tone) and the output (bilevel) image. A novel kind of massively parallel analog network (the differential neural network) is introduced and shown to be appropriate for this task. This kind of network contains a nonmonotonic nonlinearity in lieu of the sigmoid function.Keywords
This publication has 8 references indexed in Scilit:
- An adaptive least squares algorithm for the efficient training of artificial neural networksIEEE Transactions on Circuits and Systems, 1989
- Neural net based nonstandard A/D conversionElectronics Letters, 1988
- Switched-capacitor neural networksElectronics Letters, 1987
- Simple 'neural' optimization networks: An A/D converter, signal decision circuit, and a linear programming circuitIEEE Transactions on Circuits and Systems, 1986
- Image processing by simulated annealingIBM Journal of Research and Development, 1985
- “Neural” computation of decisions in optimization problemsBiological Cybernetics, 1985
- A survey of techniques for the display of continuous tone pictures on bilevel displaysComputer Graphics and Image Processing, 1976
- The effects of a visual fidelity criterion of the encoding of imagesIEEE Transactions on Information Theory, 1974