AUTOMATED VEHICLE COUNTING AND SPEED ESTIMATION USING YOLOV8 AND COMPUTER VISION

Abstract

Rapid urbanization requires very effective roadway planning to overcome traffic congestion problems. The traffic density can be monitored with manual counting or advanced instruments. However, these conventional techniques are complicated, costly, and often inaccurate. On the other hand, automized methods using video data and AI techniques provide a cost-effective method that can measure vehicle counts and speed with adequate accuracy. This paper provides a Python program for vehicle counting from recorded videos using the YOLOv8 algorithm and Computer Vision (CV). The program has been tested for three recorded videos from three different roads at different times of the day. The manual counts of the number of vehicles were recorded to assess the program's efficiency. Results showed that the program achieved an accuracy rate of 96% compared to the manual counts. This flexible program can be modified to work with streaming videos, providing real-time vehicle counting and enabling traffic congestion prediction.