Remote human sensing and conditioning of data

A disturbance is transmitted in a medium from one point to another by particle motion. The particles of the medium vibrate to transmit this energy from one particle to another. Persons moving over ground can be detected from vibrations induced to soil in the form of seismic waves, which are measured by geophones or accelerometers. Walking styles (standard, soft, or stealthy) and the background noise floor limit the detection range of footsteps. Walking style changes the dynamic footstep force on the ground and influences the footstep detection range.Seismic sensors are capable of measuring pedestrian activity and are often employed for this task for a number of reasons. Such sensor systems are inexpensive, passive (do not dissipate energy), and potentially easily installed. In this thesis a system has been developed to alarm only when a pedestrian is present. Processing of the seismic signals should make it possible for the system to discriminate footsteps from other seismic sources such as animals, railroads, operating machinery, which is a challenging problem.In this study, an algorithm has been developed to detect footsteps for security applications.Key words: footstep detection, seismic signals, intrusion detection, perimeter security