A statistical learning approach to color demosaicing - Computer Science > Computer Vision and Pattern RecognitionReport as inadecuate




A statistical learning approach to color demosaicing - Computer Science > Computer Vision and Pattern Recognition - Download this document for free, or read online. Document in PDF available to download.

Abstract: A statistical learning-inference framework for color demosaicing ispresented. We start with simplistic assumptions about color constancy, andrecast color demosaicing as a blind linear inverse problem: color parameterizesthe unknown kernel, while brightness takes on the role of a latent variable. Anexpectation-maximization algorithm naturally suggests itself for the estimationof them both. Then, as we gradually broaden the family of hypothesis wherecolor is learned, we let our demosaicing behave adaptively, in a manner thatreflects our prior knowledge about the statistics of color images. We show thatwe can incorporate realistic, learned priors without essentially changing thecomplexity of the simple expectation-maximization algorithm we started with.



Author: J.H. Oaknin

Source: https://arxiv.org/







Related documents