• Umesh Kumar

Reliable environmental perception is essential for the safe operation of autonomous and intelligent transportation systems, particularly under adverse weather conditions where visibility is severely reduced. This study presents a weather-robust object detection framework based on the YOLOv5 architecture, trained and evaluated using the DAWN dataset. The proposed approach integrates advanced preprocessing techniques, including normalization, resizing, and weather-based data augmentation (synthetic fog, rain, and low-light scenarios), to enhance detection reliability. The YOLOv5 model is further adapted with feature optimization strategies to improve object recognition under poor visibility. Performance is assessed using standard object detection metrics such as mean Average Precision (mAP) across multiple Intersection over Union (IoU) thresholds, with evaluations conducted under diverse weather conditions. Experimental results demonstrate improved detection accuracy and robustness compared to conventional methods, ensuring safer navigation of vehicles and pedestrians in intelligent transportation environments. This research contributes toward enhancing the perception capabilities of autonomous vehicles, ultimately improving road safety and traffic management in challenging weather scenarios.

Autonomous Vehicles (AVs), Object Detection, Adverse Weather Conditions, YOLOv5, Deep Learning (DL)

"Deep Learning-Based Object Detection in Smart Vehicles: YOLO for Adverse Weather", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 9, page no.d5-d11, September-2025, Available :http://www.jetir.org/papers/JETIR2509302.pdf

"Deep Learning-Based Object Detection in Smart Vehicles: YOLO for Adverse Weather", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 9, page no. ppd5-d11, September-2025, Available at : http://www.jetir.org/papers/JETIR2509302.pdf