The authors installed their prototype system over the stairs in the building and collected sensor signals when 1,593 persons are passing. The proposed algorithm shows 99% recognition
accuracy of 2-way moving direction and 95% recognition accuracy of number of passerby. Wahl et al. developed a people count system for an office space based on selfsustaining ultra-low power sensor nodes composed of a pair of uni-directional PIR sensors, an MSP430 family processor, and solar harvesting unit