Trident Dehazing Network

Most existing dehazing methods are not robust to nonhomogeneous haze. Meanwhile, the information of dense haze region is usually unknown and hard to estimate, leading to blurry in dehaze result for those regions. Focusing on these two issues, we propose a novel coarse-to-fine model, namely Trident Dehazing Network (TDN), to learn the hazy to hazy- free image mapping with automatic haze density recognition. In detail, TDN is composed of three sub-nets: the EncoderDecoder Net (EDN) is the main net of TDN to reconstruct the coarse hazy-free feature; the Detail Refinement sub-Net (DRN) helps to refine the high frequency details that was easily lost in the pooling layers in the encoder; and the Haze Density Map Generation sub-Net (HDMGN) can automatically distinguish the thick haze region with thin one, to prevent over-dehazing or under-dehazing in regions of different haze density. Moreover, we propose a frequency domain loss function to make supervision of different frequency band more uniform. Extensive experimental results on synthetic and real datasets demonstrate that our proposed TDN outperforms the state-of-the-arts with better fidelity and perceptual, generalizing well on both dense haze and nonhomogeneous haze scene. Our method won the first place in NTIRE2020 nonhomogeneous dehazing challenge.

1. Introduction

Haze is one primary source of image degradation, which results in the low contrast, color distortion and blurring problems. Previous works [21], [24] have established that haze heavily affects the performance of the high-level tasks, such as classification or semantic segmentation, and thus a preprocessing that recovers back the depicted subjects is highly desirable. To remove haze and improve visibility of the hazy image, many dehazing methods [17], [37], [29], [10], [26], [11], [20], [14], [21], [23], [27], [35] have been proposed.