Traffic Light Detection

Traffic Light Detection is a computer vision project developed to detect and classify traffic lights in real time using the YOLOv8 object detection model. The system identifies traffic light states including Red, Yellow, and Green from images, video files, or live camera streams.

The model is trained using the CinTA_v2 dataset and implemented with PyTorch and Python. It processes input frames, detects traffic lights, classifies their current state, and displays bounding boxes with labels around the detected objects. This project demonstrates how modern deep learning models can be applied to traffic monitoring and intelligent transportation systems.

The system workflow begins by taking an input video or camera stream, passing each frame through the YOLOv8 detection model, classifying the detected traffic lights, and displaying the results with bounding boxes and labels on the output stream.

This project can be applied in multiple areas such as smart traffic management, traffic monitoring systems, and autonomous driving research.

Technologies used in this project include Python, YOLOv8 (Ultralytics), PyTorch, and the CinTA_v2 Traffic Light Dataset.

Future improvements may include real-time deployment using live traffic cameras, expanding detection to include additional traffic signs and road objects, integrating the system into autonomous driving pipelines, and deploying it as a web or mobile application.