Work

Method for foggy scene restoration based on polarization information enhancement and residual total variation denoising

Year: 2025

Type: article

Abstract: Polarization-based dehazing methods can effectively address the problem of image degradation in foggy environments and significantly improve the quality of haze images. However, due to the complex and dynamic characteristics of foggy scenes, the restored images often suffer from the loss of local target detail information. To solve this problem, this paper proposes a foggy scene restoration method based on polarization information enhancement and residual total variation denoising. To fully utilize polarization information and obtain polarization information images with high information content and low noise, a normalized enhancement method with boundary threshold restriction is proposed based on statistical analysis of the pixel value distribution of numerous polarization degree images. Then, the enhanced images are subjected to residual reflection total variation regularization denoising processing. The enhanced and denoised polarization information is fused with the light intensity information using an entropy-weighted mechanism to obtain the final image restoration result. Experimental results show that the proposed method can effectively improve the detail recovery ability of polarization dehazed images.

Source: Optics Continuum

DOI: 10.1364/optcon.582147

Authors: Wenhong Gao, Zhiguo Fan, Haihong Jin, Ke Zhao, Chen Shen

Institution: Hefei University of Technology

Language: English

Topic: Image Enhancement Techniques

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

Open Access status: gold

Estimated APC paid: $1,372

Funders: National Natural Science Foundation of China, University Natural Science Research Project of Anhui Province