A Light Dual-Task Neural Network for Haze Removal

Single-image dehazing is a challenging problem due to its ill-posed nature. Existing methods rely on a suboptimal two-step approach, where an intermediate product like a depth map is estimated, based on which the haze-free image is subsequently generated using an artificial prior formula. In this pa...

Full description

Saved in:
Bibliographic Details
Published inIEEE signal processing letters Vol. 25; no. 8; pp. 1231 - 1235
Main Authors Zhang, Yu, Wang, Xinchao, Bi, Xiaojun, Tao, Dacheng
Format Journal Article
LanguageEnglish
Published IEEE 01.08.2018
Subjects
Atmospheric modeling
Dehazing
dual-task learning
Estimation
Image color analysis
Image restoration
Scattering
Task analysis
Online AccessGet full text

Cover

More Information
Summary:Single-image dehazing is a challenging problem due to its ill-posed nature. Existing methods rely on a suboptimal two-step approach, where an intermediate product like a depth map is estimated, based on which the haze-free image is subsequently generated using an artificial prior formula. In this paper, we propose a light dual-task Neural Network called LDTNet that restores the haze-free image in one shot. We use transmission map estimation as an auxiliary task to assist the main task, haze removal, in feature extraction and to enhance the generalization of the network. In LDTNet, the haze-free image and the transmission map are produced simultaneously. As a result, the artificial prior is reduced to the smallest extent. Extensive experiments demonstrate that our algorithm achieves superior performance against the state-of-the-art methods on both synthetic and real-world images.
ISSN:1070-9908
1558-2361
DOI:10.1109/LSP.2018.2849681