Abstract
Single haze removal is a challenging ill-posed problem. Most existing methods solving this dilemma depend on atmospheric physical scattering model. In other words, they recover haze-free images by estimating the atmospheric transmission. In this paper, we proposed a new recovery model called Residual Adding model, which takes dehazing procedure as a hazy image adding a loss image. Based on this new model, we proposed a single image dehazing network built with Conditional Generative Adversarial Nets (CGAN), called Residual Learning Dehazing Network (RLD-Net). Benefiting from the new model, the RLD-Net is designed as not only an end-to-end dehazing network but also a point-to-point mapping network. That means RLD-Net can take a hazy image as input and a corresponding clear image as output without any extract calculation like inversing atmospheric physical scattering model. Experimental results on both synthesized hazy images and real-world hazy images demonstrate our outstanding performance.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
