A Real-Time Deep Learning Based Unwanted Plant Mapping and Precision Spraying

Agriculture sector is the vast source of living in agricultural countries. It is suffering from challenges every year, like rapid growth in the human population, changes in climate, costly pesticides, labor availability, and price hikes. Weeds have a major effect on the worldwide agricultural economy and raise production costs by causing crop waste. Smart farming is one of the modern techniques to get high-yield from crops. For this purpose, we designed a smart robot that precisely spray over the weeds in a feasible environment. The proposed model gets <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 1. 5 \%}$</tex> accuracy to detects weeds with the raspberry pi interfaced camera using deep learning technique, such as YOLOv8. The internet of things (IoT) based smart farming system (SFS) is considered the latest technique to observe the feasible time for spray by taking real-time data of the crop. Temperature and humidity sensors are attached to the micro-controller that transmitting data to the android app to make the time-to-time log for the farmers. This approach performs the elementary function (precision spray) for SFS.