Skip to Main Content
Passenger counting is an important task for intelligent public transportation system¿¿s planning and operation. However, traditional counting systems can¿¿t identify individual passengers. The access to this additional capacity would allow collecting historical statistics on each passenger¿¿s transportation needs, hence enabling a substantially enhanced optimization in these systems. With this purpose, this work proposes a passenger counting and identification method through radio frequency identification technologies (RFID). Forty individuals are involved in measurements carrying EPC Gen2 tags in their backpacks, wallets, pocket and hand, passing through a portal with a similar size as a bus door. Conditions of one and two rows of persons walking through the portal are evaluated. This portal is equipped with 4 reading antennas. Cards that have a single EPC Gen2 tag and cards with both a Mifare and a UHF EPC Gen2 tag are used. Card reading statistics for all the combinations of 1, 2, 3 and 4 antennas are presented in this paper. The passenger detection percentage at the portal with two properly located antennas has an average of 82% for dual-band cards and 91% for single-band cards. When four antennas are used, there is an additional increase of 2% in the detection percentage for both types of cards.