COLOR PALETTE RESTORATION

To build inexpensive image display hardware, designers severely restri ct the number of colors which can be displayed simultaneously on a dev ice, A color image of a natural scene typically has tens or hundreds o f thousands of different colors in it, To display such an image on a r educed palette display device requires the image to be mapped from the full resolution color space to the reduced color space allowed by the device. Over the years, many algorithms for both selecting the reduce d palette and for mapping the full resolution color space image to the reduced palette have been developed, This mapping is lossy and degrad es the original image information. If the reduced color space image is processed using common image-processing algorithms, or if the image i s viewed on a device with a higher resolution color space, this degrad ation is very noticeable. In these cases, much better results can be o btained by first reconstructing a full color image from the reduced pa lette image. This creates a need for a palette restoration algorithm, This paper develops an algorithm for reconstructing high-color-resolut ion image data from reduced color palette images, The algorithm is bas ed on stochastic regularization using a non-Gaussian Markov random fie ld model for the image data. This results in a constrained optimizatio n algorithm which is solved using an iterative constrained gradient de scent computational algorithm, Results of the proposed palette restora tion algorithm have indicated that it is effective for the reconstruct ion of palettized images. Visual results of several experiments are pr esented. (C) 1995 Academic Press, Inc.