A color interpolation algorithm for Bayer pattern digital cameras based on green components and color difference space

Abstract

The color interpolation method of Bayer CFA used in most digital cameras plays an important role in imaging system. In Bayer pattern, the green components are the twice as much as red or blue and contain more image information, which can be used to compensate the interpolations of red and blue. Due to the high correlation among the three color channels, color mutation is prevented by the fixed color difference. Based on these principles, the proposed interpolation algorithm estimates green component firstly with adapted color plane interpolation algorithm, then computes the color difference image (R-G and B-G) to interpolate the full resolution difference image with the compensation of interpolated G and edge adaptive method. Edge blurring problem is addressed by using edge judgment to estimate RGB. Color distortion is reduced using color difference interpolation procedure. Experimental results show that the proposed algorithm outperforms existing typical interpolation methods in terms of image quality and peak signal-to-noise ratio (PSNR), and also computational complexity is low.