Adaptive reconstruction of subsampled colour images

Subsampling chrominance signals is a common practice used in colour imaging systems to reduce the storage and transmission requirements. The lossy nature of the process, however, inevitably introduces colour artifacts. In this paper, an adaptive learning system is developed for the reconstruction of full resolution colour images from subsampled signals. Simulation results are presented which show that the images produced by the new technique have fewer visible colour artifacts than those constructed by traditional techniques