Enhanced Pix2pix Dehazing Network

TLDR

The paper proposes the Enhanced Pix2pix Dehazing Network (EPDN) to transform hazy images into haze‑free ones without using a physical scattering model. EPDN consists of a GAN whose discriminator first encourages coarse‑scale realism, followed by an enhancer with two receptive‑field blocks that refine color and detail, and the GAN and enhancer are jointly trained. Extensive experiments on synthetic and real datasets show that EPDN outperforms state‑of‑the‑art methods in PSNR, SSIM, PI, and subjective visual quality.

Abstract

In this paper, we reduce the image dehazing problem to an image-to-image translation problem, and propose Enhanced Pix2pix Dehazing Network (EPDN), which generates a haze-free image without relying on the physical scattering model. EPDN is embedded by a generative adversarial network, which is followed by a well-designed enhancer. Inspired by visual perception global-first theory, the discriminator guides the generator to create a pseudo realistic image on a coarse scale, while the enhancer following the generator is required to produce a realistic dehazing image on the fine scale. The enhancer contains two enhancing blocks based on the receptive field model, which reinforces the dehazing effect in both color and details. The embedded GAN is jointly trained with the enhancer. Extensive experiment results on synthetic datasets and real-world datasets show that the proposed EPDN is superior to the state-of-the-art methods in terms of PSNR, SSIM, PI, and subjective visual effect.