Smart IoT Based People Counting System

People counting is of interest in many commercial scenarios. The number of people entering and leaving shops, the occupancy of office buildings or the passenger count of trains provide useful information to shop owners, security officials, train operators, tourism management, transport management and disaster management. To that end, this paper proposes a scheme for counting people based on a variety of approaches. One with RaspberryPi and USB webcam and another with Arduino UNO and IR sensors and further compare their accuracies. In the work it is observed that people counter using IR sensors is much more accurate than the counter which uses USB webcam and OpenCV algorithm. But different ways to increase the frame per seconds of the webcam also make RaspberryPi more efficient to process the OpenCV algorithms accurately.