Publication Date: 2023/06/13
Abstract: Agriculture plays a crucial role in the economy, and farmers strive to increase their crop yields annually. Hence effective reconnaissance is vital for farmlands and rural terrains to prevent unauthorized access and protect crops from animal damage. The expansion of agricultural lands into wildlife territories has escalated human-wildlife conflicts, with crop destruction by animals becoming a major concern. To address this, our project proposes an alerting system using YOLOv3, a real- time object detection algorithm based on deep convolutional neural networks, to classify and monitor animals that intrude into agricultural areas. This algorithm enables efficient iden-tification and tracking of animals, aiding in mitigating crop damage and ensuring the preservation of wildlife in their natural habitats. Whenever an animal is detected, the system will sendan SMS to the landowner and forest officials, providing them with early warning notifications to take appropriate actions based on the intruder’s type. This proposed system offers significant benefits to farmers, helping them increase yields and protect both humans and livestock from wild animal attacks.
Keywords: Image Processing, YOLO Algorithm, Raspberry Pi, Convolutional Neural Network, GSM Module.
DOI: https://doi.org/10.5281/zenodo.8031673
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT23MAY1948.pdf
REFERENCES