In this article, we suggest a different method for addressing the issue of nighttime single image dehazing. Because landscape frequently includes several light sources, ambient lighting haze period is usually not globally isotropic. Existing dehazing algorithms have tried to treat these two zones using same prior assumptions. We propose novel blending approach resolving them in work. A channel ...

Single-image haze removal is a challenging ill-posed problem. Recently, methods based on training synthetic data have achieved good dehazing results. However, we note that these can be further improved. A novel deep learning-based method proposed to obtain better-dehazed result for single-image in this paper. Specially, propose dual multi-scale network learn the knowledge from synthetical data....

With the development of convolutional neural networks, hundreds deep learning based dehazing methods have been proposed. In this paper, we provide a comprehensive survey on supervised, semi-supervised, and unsupervised single image dehazing. We first discuss physical model, datasets, network modules, loss functions, evaluation metrics that are commonly used. Then, main contributions various alg...

Images captured under outdoor scenes usually suffer from low contrast and limited visibility due to suspended atmospheric particles, which directly affects the quality of photos. Despite numerous image dehazing methods have been proposed, effective hazy image restoration remains a challenging problem. Existing learning-based methods usually predict the medium transmission by Convolutional Neura...

Haze removal is important for computational photography and computer vision applications. However, most of the existing methods for dehazing are designed for daytime images, and cannot always work well in the nighttime. Different from the imaging conditions in the daytime, images captured in nighttime haze condition may suffer from non-uniform illumination due to artificial light sources, which...

We introduce an improved single image haze removal algorithm, which combines dark channel prior (DCP) and histogram specification. First, the dark channel prior knowledge proposed by Kaiming He is analyzed and a conclusion is drawn that the haze removal image based on dark channel prior will have a tendency to dim and indistinct in some specific situations. Especially, when cleaning the haze in...

In this paper, we study two challenging and less-touched problems in single image dehazing, namely, how to make deep learning achieve dehazing without training on the ground-truth clean (unsupervised) an collection (untrained). An unsupervised model will avoid intensive labor of collecting hazy-clean pairs, untrained is a “real” approach which could remove haze based observed hazy only no extra...