Image demosaicing using content and colour-correlation analysis

•We propose a content adaptive demosaicing strategy utilising structure analysis and correlation between the red, green and blue planes.•The filter coefficients are optimised through an offline training procedure.•The online filtering process is very efficient and performs comparatively with much more expensive methods.

Most digital cameras of today use a single CCD image sensor with alternating arrangement of red, blue and green colour filters in what is known as a Bayer pattern. To extract a full colour image, a demosaicing strategy has to be applied. In this paper we propose a content adaptive demosaicing strategy utilising structure analysis and correlation between the red, green and blue planes. These two aspects are used for the classification of a block of pixels to generated trained filters. The proposed method aims to reconstruct a high quality demosaiced image from a Bayer pattern in a colour filter array efficiently. Experimental results show that the proposed strategy performs comparatively as more expensive methods.