Efficient demosaicking of color images using theoretic reconstruction technique

Demosaicking is the process of obtaining a RGB (Red, Green and Blue) Image from the Color filter Array (CFA) applied color Image . In general a good demosaicking algorithm reduces the low frequency information of imagery, since such information is not essential for image visual quality. This paper proposes an influential demosaicking technique that uses constraint sets to reconstruct a full color resolution channel for Bayer CFA samples. The proposed algorithm produces high frequency information of imagery such as possible. This iterative set theoretic reconstruction technique converges to a fixed point which is in fact solution to the demosaicking problem. Experimental results of the proposed method is much better than compare to other latest demosaicking algorithms in terms of image quality and Peak signal to Noise ratio (PSNR).