A Single Image Dehazing Technique Using the Dual Transmission Maps Strategy and Gradient-Domain Guided Image Filtering

Syed Muhammad Ehsan,Anayat Ullah,Ersin Elbasi,Muhammad Imran

doi:10.1109/access.2021.3090078

Abstract

In this paper, a single image dehazing technique using dual transmission maps strategy and gradient-domain guided image filtering is presented. A new strategy is adopted to compute the dual transmission maps using the dark channel and atmospheric light. Further, the transmission maps are refined to remove any remaining ill effects using the gradient-domain-guided filter. Finally, using the dark channel, atmospheric light, and refined transmission map, the haze-free image is obtained. The dual transmission maps strategy not only removes halo artifacts and reduces the saturation but also ensures the natural appearance in the recovered images. Furthermore, the proposed scheme is evaluated using a wide range of images and compared with state-of-the-art schemes. The comparison shows the superiority of the proposed technique in terms of recovering haze-free images.

Highlights

Adverse weather, when taken outdoor images, could often decline the quality of images significantly
The proposed technique utilizes three parameters: atmospheric dispersion model, dark channel prior, and transmission map; these parameters along with other information are discussed in detail
ATMOSPHERIC DISPERSION MODEL The atmospheric dispersion model is a physical model used for the formation of haze [1]

Summary

INTRODUCTION

Adverse weather, when taken outdoor images, could often decline the quality of images significantly. Most of the techniques, discussed so far, experience hindrances like computational complexity and fail to achieve good results for images containing large sky regions To solve this problem of sky failure, a very effective and efficient method has been proposed in [23]. The method proposed in [24] is a very fast and efficient dehazing single image algorithm that uses a dark channel and morphological reconstruction processes to refine the transmission map and avoid computational complexity. Neural network-based dehazing methods give good results, but, like other neural-network-based techniques, they need large datasets and high computational resources These requirements make them unsuitable for real-life applications.

BACKGROUND

EXPERIMENTAL RESULT

CONCLUSION