A Deep Learning Based Model for Detection and Tracking of Stray Animals in the Fields

Nitika Goyal

Department of Computer Science, Guru Nanak College, Budhlada, Punjab, India.

Corresponding Author E-mail:goyalniti82@gmail.com

DOI : http://dx.doi.org/10.12944/CARJ.12.3.38

Abstract:

Agriculture, despite of being a major contributor in Indian economy, has to confront innumerable issues like intrusion of stray animal in agricultural areas, giving rise to economical loss due to damage of crops. The inefficiency of conventional practices like trenches, fences etc in checking animal intrusion has paved way for technology to provide innovative solutions in the field of agriculture. An innovative approach integrating deep learning model YOLOv5 with IoT devices has been proposed in this paper for timely detection and tracking of entry of stray animals in the fields and giving instant alert to the farmer about the intrusion in order to protect crops from damage and prevent human-animal conflict. The proposed system had been tested on numerous classes of animals and exceptional accuracy of above 95% could be achieved. Furthermore the alert system not only gives timely notification about the intrusion but also specify sufficient details about the intruding animal. However, challenges like high initial setup cost, availability of high speed internet connection, and handling diverse animal species in different geographical areas can create hindrance in implementation of this model in developing nations. The research work highlights the capability of technological advancements in tackling critical issues in the field of agriculture by providing innovative solutions specifically in case of crop damage by animal intrusion. Future work can focus on adaptation of the model for diverse applications like Biosphere reserve management, forest management etc.

Keywords:

Animal intrusion; Crop protection; IoT; Real-time alert; YOLOv5

Download this article as: 

[ HTML Full Text]