A Fast Image Dehazing Algorithm for Highway Tunnel Based on Artificial Multi-exposure Image Fusion. Issue 1 (January 2020)

• Record Type: Journal Article
• Authors: Li, Wenfeng; Wei, Hongyan; Qi, Guanqiu; Ding, Hao; Li, Ke
• Is Part Of: IOP conference series. Volume 741:Issue 1(2020)
• Publication Date: 2020-01
• Subjects: Materials science -- Periodicals; 620.1105
• DOI: 10.1088/1757-899X/741/1/012038
• Languages: English
• ISSNs: 1757-8981

Abstract

Abstract: Single image dehazing has been a challenging problem due to its ill-posed nature. This paper proposed a fast image dehazing algorithm for highway tunnel based on artificial multi-exposure image fusion. The original blurred images are transformed into under-exposed images using a series of gamma-correction operations. Meanwhile, a multi-exposure image fusion algorithm based on two-scale decomposition is proposed. It exploits the guided filter to separate each multi-exposure image into base and detail layers. The fused base and detail layers are integrated into the fused images. Then, a linear saturation adjustment method is utilized to enhance the saturation in spatial domain. Through the above image processing steps, the final haze-free images can be obtained. In this paper, the superiority of the proposed image dehazing algorithm based on artificial multi-exposure image fusion is demonstrated through theoretical analysis and experimental verification. It can quickly improve the visibility of foggy images, and make the processed images colorful and clear.