YOLOv4-Based Object Recognition Algorithm for Traffic Monitoring

Intelligent transportation systems currently require reliable, real-time vehicle detection from visual and audio data for traffic monitoring, and these activities have become crucial in recent years. Machine learning is one of the most important technologies to address this issue since it allows for the perception of information about the environment around the vehicle, which is vital for safe driving. In this study we have implemented the upgraded YOLOv4 video stream object detection algorithm in combination with virtual detector, blob tracking to analyse the video footage of the traffic flow recorded by a camera. Also, we have applied Open CV Computer Vision library to detect objects from the image, track, count and classify the moving vehicles.