English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This paper presents a comprehensive study of applying the convolutionalneural network (CNN) to solving the demosaicing problem. The paper presents twoCNN models that learn end-to-end mappings between the mosaic samples and theoriginal image patches with full information. In the case the Bayer colorfilter array (CFA) is used, an evaluation with ten competitive methods onpopular benchmarks confirms that the data-driven, automatically learnedfeatures by the CNN models are very effective. Experiments show that theproposed CNN models can perform equally well in both the sRGB space and thelinear space. It is also demonstrated that the CNN model can perform jointdenoising and demosaicing. The CNN model is very flexible and can be easilyadopted for demosaicing with any CFA design. We train CNN models fordemosaicing with three different CFAs and obtain better results than existingmethods. With the great flexibility to be coupled with any CFA, we present thefirst data-driven joint optimization of the CFA design and the demosaicingmethod using CNN. Experiments show that the combination of the automaticallydiscovered CFA pattern and the automatically devised demosaicing methodsignificantly outperforms the current best demosaicing results. Visualcomparisons confirm that the proposed methods reduce more visual artifacts thanexisting methods. Finally, we show that the CNN model is also effective for themore general demosaicing problem with spatially varying exposure and color andcan be used for taking images of higher dynamic ranges with a single shot. Theproposed models and the thorough experiments together demonstrate that CNN isan effective and versatile tool for solving the demosaicing problem.

Learning Deep Convolutional Networks for Demosaicing