PIR Sensor Based Elephant Intrusion Detection System Using Faster R-CNN Algorithm

Authors

  • Jothibasu M. , Harsha R. , Naveen K.S. , Suriyaprakash T.B. , Usha S

Abstract

The increasing fragmentation of forest lands and isolation of elephant populations
becomes the main cause of Human Elephant Collision. Human-Elephant conflict (HEC) is a
major problem in the agriculture field and in the forest zone leading to crop damage and put
human life in danger. The rapid detection is necessary to prevent loss and could be achieved
through the proposed Elephant Intrusion Detector based on vision based Surveillance (EIDS)
that alerts the residents and forest rangers, thus reducing the Human Elephant Collision. This
paper briefly explains the application of PIR sensor in the intrusion detection system and the
implementation of Faster R- CNN algorithm for the detection of elephants from surveillance
camera video. The PIR sensors will detect the movement of the object in the range of (1 - 8)
m and it trigger the camera to record a video for 10 sec. The captured video is converted as
images at the rate of single frame/sec and it is processed by Faster R-CNN algorithm to
examine the existence of the elephant. If the elephant is identified, warning message will be
sent to forest officials indicating that an elephant has been detected and is approaching
human habitat, using GSM module. This prototype/model provides a smarter solution that
could protect the human/crops from animal intrusions without causing any harm to the
wildlife.

Published

2020-12-01

Issue

Section

Articles