Fog Detection through Image Processing Methods

Abstract

This paper presents a fog detection algorithm, highlighting the significance of continued exploration in fog identification through image processing techniques. The advancement and application of this algorithm can significantly benefit various domains, including road safety, environmental monitoring, navigation, security, surveillance, and improving existing systems’ performance. The evaluation performed on test images have shown an accuracy of 72%, a precision of 94%, a recall of 57% and an F1 score of 0.71. The proposed algorithm clearly outperformed some existing fog detection methods.