研究目的
To develop a robust framework for enhancing the visibility of images degraded by foggy weather conditions by combining a modified trilateral filter-based visibility restoration approach with an S-shaped transfer function for contrast enhancement.
研究成果
The proposed defogged algorithm effectively enhances visibility in foggy images by combining trilateral filtering and S-shaped transfer function, outperforming existing methods in terms of FRF, EMF, and visual quality, making it suitable for applications like image analysis and weather forecasting.
研究不足
The algorithm may not handle all types of weather degradations beyond fog, and computational complexity could be a concern for real-time applications. The study is limited to specific image databases and may not generalize to all foggy conditions.
1：Experimental Design and Method Selection:
The proposed algorithm involves two phases: first, a modified trilateral filter is used for visibility restoration to smooth and defog images; second, an S-shaped transfer function is applied for contrast enhancement. The method is based on the Koschmieder model for image degradation in fog.
2：Sample Selection and Data Sources:
Test images are sourced from Tarel et al. foggy image database and Lark Kwon Choi et al. foggy image database, including formats like PNG, BMP, and JPEG, with specific images such as y11_photo.png, cones.jpg, etc.
3：List of Experimental Equipment and Materials:
A 64-bit system with Core i3 processor, 4 GB RAM, and MATLAB [release 2015a] software is used for implementation.
4：Experimental Procedures and Operational Workflow:
The algorithm is applied to foggy images in two steps: trilateral filtering for defogging followed by S-shaped mapping for contrast enhancement. Performance is evaluated using metrics like FRF, EMF, and g.
5：Data Analysis Methods:
Quantitative analysis is performed using Fog Reduction Factor (FRF), Measure of Enhancement Factor (EMF), and blind parameter (g). Visual results are compared, and mean opinion scores (MOS) are collected from experts.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
