SmartCrop: An Intelligent IoT System for Automated Weed Detection and Precision Pesticide Spray

The rapid advancements in IoT, image processing, and Machine Learning (ML) have presented numerous opportunities for diverse applications. One such application is the classification of images for day-to-day agricultural tasks. Effective weed monitoring and control are crucial in agriculture, necessitating the use of pesticides. This paper introduces a novel hardware-based implementation using Raspberry Pi, presenting the design of a robot that automates the process of weed detection and pesticide application. The proposed work utilizes the makesense.ai software tool to label input images and assign corresponding names, while leveraging the Yolov5 architecture, a Deep Learning (DL) based object detector, for image segregation in our system. The obtained results demonstrate efficient weed detection and control, contributing to farmer safety by reducing pesticide exposure, saving time, and minimizing manual labor. This research significantly advances agricultural practices, offering a solution that enhances productivity and promotes a safer and more sustainable farming approach.