An improved color image demosaicking algorithm
Shousheng Luo, Haomin Zhou, Tie Zhou · 2012
Color image demosaicking is an important issue in color image acquisition procedures. Most of the existing demosaicking methods reduce the artifacts at the price of the high computational cost posterior processing. In this paper, we propose a more effective demosaicking algorithm. The proposed method improves the qualities of demosaicked images by interpolation, rather than posterior processing. In order to avoid the propagation of errors occurred during green channel reconstruction, we interpolate red and blue channels by original samples to the greatest extent. Experimental results show that the proposed algorithm yields better demosaicking images as compared with a number of typical demosaicking algorithms. Furthermore, the computational cost of the proposed method is lower.