Deep Residual Network For Msfa Raw Image Denoising

Multispectral filter arrays (MSFA) is increasingly used in multispectral imaging. Although many previous works stud- ied the denoising algorithms for color filter array (CFA) based cameras, denoising MSFA raw images is little discussed. Compared to CFA, the major challenges for denoising MSFA data include 1) MSFA data contains more channels than CFA and no channel is predominant; 2) it is non-trivial to design the denoising process to be compatible with the subsequent demosaicking process. To overcome these challenges, we propose a new deep residual network that is dedicated to the MSFA mosaic patterns. First, a split and stride convo- lution layer is introduced to match the mosaic pattern of the MSFA raw image. Second, we apply a novel data augmen- tation using MSFA shifting and dynamic noise to make the model robust to different noise levels. Third, a new network optimization criteria is created by using the noise standard deviation to normalize the L1 loss function. Comprehensive experiments demonstrate that the proposed deep residual net- work outperforms the state-of-the-art denoising algorithms in MSFA field.