A Fast Method Of Fog And Haze Removal

2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)(2016)

引用 13|浏览4
暂无评分
摘要
Fog and haze degrade the quality of preview and captured image by reducing the contrast and saturation. As a result the visibility of scene or object degrades. The objective of the present work is to enhance the visibility, saturation, contrast and reduce the noise in a foggy image. We propose a method that uses single frame for enhancing foggy images using multilevel transmission map. The method is fast and free from noise or artifacts that generally arise in such enhancement techniques. A comparison with existing methods shows that the proposed method performs better in terms of both processing time and quality. The proposed method works in real time for VGA resolution. The proposed work also presents a scheme to remove fog, rain and snow in real-time.
更多
查看译文
关键词
Fog removal,snow removal,rain removal,dark channel prior,temporal filter
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要