Search

Bhutan primarily relies on land transportation; hence, diverse technologies are crucial for ensuring the safety of the drivers. Dust, fog, smoke, and other dry particles often obstructs the road visibility. Furthermore; complex mountainous terrain and steep topography increase fog formation, while winding winds heighten accident risks. This study aims to reduce weather-related accidents by introducing an IoT-centric system designed to enhance visibility and provide timely alerts to drivers. The system employs the Dark Channel Prior (DCP) algorithm for real-time foggy video processing and refining of dehazed video using the Guided Filtering method. This research also focuses on developing an alert system using a time-of-flight (ToF) sensor with detection range of 15 m. Real-time testing and validation were conducted along the fog-prone Thimphu-Phuentsholing highway, specifically in the Jumja region. To evaluate the performance of the system; validation metrics, including the peak signal-to-noise ratio (PSNR) and Structural Similarity Index Measure (SSIM) were determined for quantitative analysis, and the original and dehazed videos were compared on the same screen for visual analysis. The effectiveness of obstacle detection system was evaluated by comparing the actual measured distance to the distance displayed on the screen. The results depicts that the proposed system significantly improves visibility and effectively sirens drivers to impending obstacles within the sufficient stopping sight distance through display of clear video in the dehazing system. The study highlights the potential of advanced IoT and computer vision technologies to enhance road safety and reduce accident rates under low-visibility conditions.