Real-Time Railway Obstacle Detection in Variable Weather Conditions: A Novel Framework for Enhanced Safety Using YOLOv8

Railway safety is a significant concern for society due to the risks posed by objects obstructing train tracks. To address this issue, various approaches have been proposed, and this paper introduces a novel solution using YOLOv8, an advanced object detection technology. YOLOv8 enables the system to effectively recognize and address potential risks on railway tracks by identifying objects ahead of a train’s locomotive, including humans, animals, and other hazardous obstacles. The objective of the paper is to reduce the frequency of railway accidents and enhance safety for all individuals involved by using YOLOv8 model with feature extraction, data augmentation, and pre-processing techniques. Particularly in rainy and foggy seasons, with high precision in obstacle detection.