Crop Protection Against Wild Animal and Alert System using Deep Learning

Abstract

A crop's vulnerability to animal assaults exists. Protecting agricultural areas is a difficult task these days. It is imperative to search for any critters that might have a more detrimental effect on the crop as a result. The following creatures can be found in the protected area: The preservation of the grain crop is of utmost importance due to its sustained attacks over time. The topic has been approached in a way that makes the current approaches ineffective; in this study, we propose a strategy to protect farms from wild animals without causing harm to them, thereby creating a system that considers their needs (deer, nilgai, wild boar, etc.). To find the animal and make a loud noise to frighten them. In order to address this issue, the suggested approach uses the Mobile Net concept to create an IOT and deep learning based model that detects the early arrival of wild animals. When a wild animal is spotted, an automatic stone gun controlled by an Arduino UNO with a relay module triggers a servo motor to launch stones at the animals without endangering them. Keywords: Priority, Mobile Net, IOT, Deep Learning, Arduino UNO.