We propose an algorithm to find the footstep classification of person movement using seismic sensor. A field study were conducted in 2011 at Jeju national university, to document seismic signature of a person who is walking crawling and running in ground. The person generated a vertical ground vibration were recorded using 2000 Hz of sampling frequency with geophone (GD20DX). The hypothetical sensor performance predictions were made using a s statistical signal processing approach is developed footstep classification of the movement of a person. We experimentally verified the algorithm of the footstep classification by using the collected seismic data.