ESTD Year: 2017 | Impact Factor (2026): 8.7
DOI Prefix: 10.47001/IRJIET
Vol 1 No 1 (2017): Volume 1, Issue 1, September 2017 | Pages: 9-13
International Research Journal of Innovations in Engineering and Technology
OPEN ACCESS | Research Article | Published Date: 03-09-2017
Digital cameras usually has a single sensor covered with a color filter array which samples only one color at location of every pixels. The full color image is restored using a demosaicing technique through means of a digital pipeline imaging. Here we propose a novel approach of demosaicing based on the directional filtering and weighting techniques. The contributions this system is doubled. First of all we analyze the limitations and advantages of the existing directional filtering and weighting techniques. Then we improve those two existing techniques in order that it suppresses all the common demosaicing artifacts. In second, we develop new estimation schemes for the color components reconstruction. Through experimentation we show the proposed results can outperform all the recent unique demosaicing algorithms in terms of accuracy and objective measures.
Color filters, Demosaicing, Directional Filtering, Weighting technique, MATLAB, RGB sensor
This work is licensed under Creative common Attribution Non Commercial 4.0 Internation Licence