Adaptive Image Demosaicing Algorithm Based On K-Nearest Neighbor for Improved Visual Quality
Gurjot Kaur Walia, Jagroop Singh · SAMRIDDHI A Journal of Physical Sciences Engineering and Technology · 2022
Demosaicing extracts a high quality, full-color image from the incomplete data samples obtained through image sensors via bayer pattern. Healthcare, image forensics, low light photos, etc. use this technique. For a major consideration, this work provides an adaptive demosaicing approach that uses gradient corrected linear interpolation along with k-Nearest Neighbor algorithm to learn from the labelled training set, with the output based on a distance measurement. Signal to Noise Ratio, Peak Signal to Noise Ratio were determined in order to justify the effort and the findings showed that the mentioned method produced better results.