Fusion of Heterogeneous Adversarial Networks for Single Image Dehazing

In this paper, we propose a novel image dehazing method. Typical deep learning models for dehazing are trained on paired synthetic indoor dataset. Therefore, these models may be effective for indoor image dehazing but less so for outdoor images. We propose a heterogeneous Generative Adversarial Networks (GAN) based method composed of a cycle-consistent Generative Adversarial Networks (CycleGAN) for producing haze-clear images and a conditional Generative Adversarial Networks (cGAN) for preserving textural details. We introduce a novel loss function in the training of the fused network to minimize GAN generated artifacts, to recover fine details, and to preserve color components. These networks are fused via a convolutional neural network (CNN) to generate dehazed image. Extensive experiments demonstrate that the proposed method significantly outperforms the state-of-the-art methods on both synthetic and real-world hazy images.

I. Introduction

Haze is defined as a condition that dust, smoke particulates, or water droplets obscure the clarity of the atmosphere. As light gets scattered or absorbed through these particles, subjects in the scene lose contrast. Haze often causes significant degradation of the image quality and it may adversely affect performance of computer vision algorithms. Additionally, these images typically exhibit degraded color information as light absorption in the medium varies based on the wavelength. As a result, foggy images often show greenish/bluish tinge with red component being suppressed. To mitigate this problem, various dehazing methods are presented.