Hybrid mean adaptive center weighted median filter for color sensor denoising
Tamer Farouk Rabie · 2005
The paper describes a novel method for removing sensor noise from images acquired using digital cameras at high ISO settings. The problem lies in the fact that the algorithm must deal with hardware-related sensor noise that affects certain color channels more than others and is thus non-uniform over all color channels. An adaptive signal equalization mean and median hybrid filter is formulated and experimental results are presented with an analysis of the performance of the filter in comparison with other standard noise reduction filters.