A Real-Time Vehicle Detection and Counting System Using a GUI Interface.

The growing need for intelligent traffic monitoring and environmental control has fuelled the importance of automated vehicle detection and counting systems. Fuel consumption estimation, vehicular emissions calculations, and traffic pattern comprehension- all play a vital role in shaping sustainable environmental policies. Therefore, accurate identification and counting of vehicles is essential. In this paper, we have shown the development of a real-time vehicle detection and counting application that uses computer vision and deep learning techniques to effectively detect, categorise, and count several types of vehicles on Indian roads. The computer vision vehicle counting application uses the YOLO (You Only Look Once) deep learning model for real-time identification and categorisation of diverse vehicle types, including cars, buses, trucks, and motorbikes [1]. A tracking module based on the SORT (Simple Online and Real-time Tracking) algorithm ensures unique identification of vehicles across frames and prevents duplicate counting [2]. When the centre of a vehicle’s bounding box crosses a predetermined virtual line, the system updates the count of that vehicle type. The system also incorporates a graphical user interface (GUI) that supports video input and real-time visualisation of results, making it user-friendly and adaptable for practical deployment. The proposed application offers valuable insights for intelligent city initiatives, traffic management, and data-driven urban planning. It can also be used to support environmental sustainability by providing detailed vehicle activity data to calculate emissions, energy consumption, and policy planning for cleaner and more efficient urban mobility.