Color image restoration using compound Gauss-Markov Random Fields

In this work we extend the use of Compound Gauss Markov Random Fields to the restoration of color images. While most of the work in color image restoration is concentrated on enforcing similarity between the intensity values of the pixels in the image bands, we propose combining information by means of the line process. In order to find the multichannel restoration modified versions of ICM and SA are proposed. The methods are finally tested on real images.