Orthogonal Transform based Generative Adversarial Network for Image Dehazing

Abstract

Image dehazing has become one of the crucial preprocessing steps for any computer vision task. Most dehazing methods work in the image domain, and the dehazed image is obtained by estimating the transmission map along with global atmospheric light. In this thesis, we present a novel end to end architecture for estimating dehazed image in the Krawtchouk transform domain. For this a customized Krawtchouk Convolution Layer (KCL) in the architecture is added. KCL is constructed using Krawtchouk basis functions which converts the image from the spatial domain to the Krawtchouk transform domain. At the end of the architecture, another convolution layer called Inverse Krawtchouk Convolution Layer (IKCL) is introduced which converts the image back to the spatial domain from the transform domain. It has been observed that the haze is primarily present in lower frequencies of hazy images. Krawtchouk transform helps to analyze the high and low frequencies of the images separately. We have divided our architecture into two branches, the upper branch deals with the higher frequencies while the lower branch deals with the lower frequencies of the image. The lower branch is made deeper in terms of the layers as compared to the upper branch to address the haze present in the lower frequencies. When compared to current state-of-the-art methods, we were able to get competitive results using the proposed Orthogonal Transform based Generative Adversarial Network (OTGAN) architecture for image dehazing.