Perceptual versus gaussian smoothing for pattern-colour separabilityM. Mirmehdi, M. Petrou, Perceptual versus gaussian smoothing for pattern-colour separability. International Conference on Signal Processing and Communications. ISBN 0-88986-246-X, pp. 136–140. February 1998. PDF, 184 Kbytes.
This paper brings to the fore a set of masks that are used in combination to simulate the pattern-colour separability function of the human visual system. These kernels give a perceptual degree of smoothing which correspond to measurements estimated from human psychophysical experiments. We have found that they are a better approach to image smoothing for colour images than the traditional Gaussian masks which are widely practised.