JJEM: Special Edition 2 (December 2024)
JJEM: Special Edition 2 (December 2024)
2024-12-08
Detection of Road Humps and potholes using yolo v3.
Hemanthraj S N, Manjunath H T
Road anomalies like potholes and speed bumps must be recognized and categorized in order to maintain road safety. This study introduces an object identification technique, You Only Look Once version 3 (YOLOv3), that provides a robust way to identify potholes and road humps. YOLOv3 is well-known for its real-time object detection speed and precision, which makes it appropriate for use in dynamic contexts like road condition monitoring. Our method entails using a bespoke dataset of photos of potholes and road bumps that have been suitably labeled for these particular features to train the YOLOv3 model. To guarantee the model's resilience in a variety of settings, the dataset contains a range of weather patterns, lighting setups, and road conditions. After training, the model is assessed on an independent test set to determine its the findings show that YOLOv3 can accurately and reliably identify between potholes and road bumps with high recall rates. Furthermore, the model's real-time detection capabilities enable a smooth integration into automotive systems, providing drivers with instantaneous notifications and alerting maintenance teams when repairs need to be made.
Algorithm for CNN, Pothole Detection, Hump Detection.