People Matching for Transportation Planning Using Texel Camera Data for Sequential Estimation

This paper addresses automatic people matching in the dynamic setting of public transportation, such as a bus, as people enter and then at some later time exit from a doorway. Matching a person entering to the same person exiting at a later time provides accurate information about individual riders, such as how long a person is on a bus and the associated stops the person uses. At a higher level, matching exits to previous entry events provides information about the distribution of traffic flow across the whole transportation system. The proposed techniques may be applied at any gateway where the flow of human traffic is to be analyzed. For the purpose of associating entry and exit events, a trellis optimization algorithm is used for sequence estimation, based on multiple texel camera measurements. Since the number of states in the trellis exponentially grows with the number of persons currently on the bus, a beam search pruning technique is employed to manage the computational and memory load. Experimental results using real texel camera measurements show 96% matching accuracy for 68 people exiting a bus in a randomized order. In a bus route simulation where a true traffic flow distribution is used to randomly draw entry and exit events for simulated riders, the proposed sequence estimation algorithm produces an estimated traffic flow distribution, which provides an excellent match to the true distribution.