Tracking and counting multiple people using distributed seismic sensors

Heel strikes from walking people generate impulses that propagate as Rayleigh waves that can be measured by seismic sensors. The time of arrival of these waves dependent upon both the location of the sensors and the propagation medium, which is usually not known. Hence, associating measured signals to an individual person is difficult. In this paper, we present two approaches for grouping the time of arrival measurements of the impulses and use them to estimate the number of targets using an iFilter implemented as a particle filter. The algorithms are used to process the actual data collected using distributed seismic sensors and their results also presented.