Weighted Nuclear Norm and TV Regularization based Image Deraining

Abstract

Often, images captured by digital camera in outdoor vision system may be significantly distorted by bad weather conditions. Such visual distortions may negatively affect the performance of the system. One such bad weather condition is rain, which randomly makes intensity fluctuations in the images. This paper proposes a new low rank recovery based algorithm to remove the rain streaks from single image taken in rainy weather. This method makes the use of weighted nuclear norm (WNN) and total variation (TV) regularization for efficient rain removal. WNN assigns different weights to different singular values based on the details each singular value holds. TV regularization is used to discriminate most of natural image content from sparse rain streaks by preserving piecewise smoothness of images. Simulation result shows that the rain streaks are more effcaciously eliminated by our method.