Digital cameras usually use a single sensor covered with a colour filter array which samples only one colour at the location of each pixel. Restoring a full-colour image by the demosaicking technique is a key task in the digital imaging pipeline. This study proposes a novel demosaicking approach based on the existing directional filtering and weighting techniques. The contributions of this study are two-fold. First, the authors analyse the advantages and limitations of the existing directional filtering and weighting techniques, and improve the two existing techniques in order to suppress the common demosaicking artefacts. Second, the authors give a new estimated scheme for the reconstruction of the colour components. The experimental results show that the proposed method outperforms the recent six state-of-the-art demosaicking methods in terms of both subjective and objective measures.
Image interpolation is a technique of producing a highresolution image from its low-resolution counterpart, which is often required in many image processing tasks. In this paper, we propose an edge-directed bicubic convolution (BC) interpolation. The proposed method can well adapt to the varying edge structures of images. The experimental results show that it reduces common artifacts such as blurring, blocking and ringing etc. and significantly outperforms some existing interpolation methods (including BC interpolation) in terms of both subjective and objective measures.
This paper reviews the NTIRE 2022 challenge on efficient single image super-resolution with focus on the proposed solutions and results. The task of the challenge was to super-resolve an input image with a magnification factor of ×4 based on pairs of low and corresponding high resolution images. The aim was to design a network for single image super-resolution that achieved improvement of efficiency measured according to several metrics including runtime, parameters, FLOPs, activations, and memory consumption while at least maintaining the PSNR of 29.00dB on DIV2K validation set. IMDN is set as the baseline for efficiency measurement. The challenge had 3 tracks including the main track (runtime), sub-track one (model complexity),
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.