Autonomous Seismic Sensor with a New Temporal Method of a Moving Person Detection

Abstract

The article presents an autonomous seismic sensor developed to detect a moving person using temporal characteristics of a seismic signal. In order to exclude false alarms associated with moving animals, an additional temporal informative classification feature of the seismic signals from a moving person was introduced into a person detection algorithm, in contrast to known seismic sensors. This feature is associated with a difference in a movement nature of humans and animals. Sensor tests have shown that it has a significant probability of the correct person detection along with low power consumption. It is shown that such characteristics can be achieved using a relatively simple temporal method of the moving person detection. The range of use of the developed sensor is the protection of various objects, border and remote monitoring of the people movement.