| CPC G06T 3/4015 (2013.01) [G06T 3/4046 (2013.01); G06T 11/001 (2013.01)] | 20 Claims |

|
1. A demosaicing method based on deep learning, comprising:
providing an artificial neural network configured to convert a mosaic image representable by pixels with color pixel values corresponding to a plurality of colors to a plurality of color difference images corresponding to the plurality of colors;
simultaneously generating a plurality of output color difference images corresponding to the plurality of colors based on an input mosaic image representable by pixels with color pixel values corresponding to the plurality of colors using the artificial neural network, wherein the plurality of output color difference images include a red color difference image configured to restore a demosaiced red image, a green color difference image configured to restore a demosaiced green image, and a blue color difference image configured to restore a demosaiced blue image; and
generating a plurality of demosaiced color images corresponding to the plurality of colors by summing, pixel by pixel, the input mosaic image and each of the plurality of output color difference images.
|