文摘
TransJakarta is a BRT-based mass transportation system operating in Jakarta, the capital of Indonesia. Currently, the system delivers rather low level of service; as a result, the number of passengers is rather low in comparison to that of the other BRT-based systems. In this work, we propose a design of the passenger counting system for BRT-based system, which is very important for the BRT fleet management to increase the system level of service. The counting system is established by deploying computer vision techniques. A few algorithms are evaluated and the Adaptive Median Filtering with the sampling rate of 13 produces the highest level of precision and recall.