Image Restoration by Matching Gradient Distributions

时 间:2012年5月29日 (周二)上午 9:30

地 点:北京大学计算机科学技术研究所 309 会议室 (中关村北大街128号计算机所大楼)

报 告 人:任杰

报告题目:Image Restoration by Matching Gradient Distributions

报告摘要:

The restoration of a blurry or noisy image is commonly performed with a MAP estimator, which maximizes a posterior probability to reconstruct a clean image from a degraded image. A MAP estimator, when used with a sparse gradient image prior, reconstructs piecewise smooth images and typically removes textures that are important for visual realism. In this paper, the authors present an alternative deconvolution method called iterative distribution reweighting (IDR) which imposes a global constraint on gradients so that a reconstructed image should have a gradient distribution similar to a reference distribution. In natural images, a reference distribution not only varies from one image to another, but also within an image depending on texture. In this work, a reference distribution is estimated directly from an input image for each texture segment. The proposed algorithm is able to restore rich mid-frequency textures. A large-scale user study supports the conclusion that the proposed algorithm improves the visual realism of reconstructed images compared to those of MAP estimators.

注:文章选自 IEEE Trans. on PAMI, VOL 34, NO. 4, April 2012, pp.683-694.

CLOSE

上一篇 下一篇