Title :
A low cost fence impact classification system with neural networks
Author_Institution :
French South African Tech. Inst. in Electron., Pretoria
Abstract :
Due to a high rate in live stock theft and farm intrusions, securing property is a problem in South Africa. Research is necessary to investigate possible improvements. By analyzing the acoustical vibrations that are generated on a game farm or security fence during entry, intrusions can be detected. Owners of protected property can receive an early warning. Different type of intrusions can be distinguished since they generate different vibrations. A point sensor is used to detect vibrations generated at a distance from the sensor. This proposed system is part of a more elaborate farm information system and is not intended to function as a stand-alone security system
Keywords :
acoustic signal processing; alarm systems; neural nets; safety systems; signal classification; vibration measurement; acoustical signal classification; farm information system; farm intrusions; fence mount sensors; game farm; intrusion detection; live stock theft; low cost fence impact classification system; neural networks; perimeter protection; point sensors; property security; security fence; vibration measurement; vibration sensors; Acoustic sensors; Acoustic signal detection; Africa; Cameras; Costs; Information security; Neural networks; Protection; Sensor systems; Statistics;
Conference_Titel :
AFRICON, 2004. 7th AFRICON Conference in Africa
Conference_Location :
Gaborone
Print_ISBN :
0-7803-8605-1
DOI :
10.1109/AFRICON.2004.1406647
