Noise insensitive demosaicing algorithm
Changwon Kim, Moon-Gi Kang, Sungdae Cho, Sangwook Oh, Hyunhee Park, Minkyu Park · IEEE International Conference on Signal and Image Processing · 2007
The problem of recovering full-color images from a noisy and color-sampled observation is considered in this paper. Instead of removing the sensor noise before or after the color interpolation step, we propose a novel technique to combine demosaicing and denoising procedures simultaneously. In order to avoid artifacts in high frequency regions and improve the performance, an edge indicator function is used and horizontal and vertical direction of edges are considered in the proposed method, while noise is simultaneously removed by adaptive denoising. As a postprocessing, the modified filteringon color difference domain is adopted to improve the quality of the image. Experimental results illustrate the benefits of the proposed method. When compared to the conventional methods, the proposed method outperforms them on quantitative and qualitative criteria.