Automated Weed Detection for Crop Health Monitoring

Abstract

Smart agriculture provides an opportunity for a step-change in agricultural productivity. When it comes to crop yield, weed management is the most important aspect. They are the most impacting factors which act as a hindrance in overall crop productivity and causes important yield loss worldwide. Use of technology in the field of agriculture can help automate the process of weed detection. This will not only be highly efficient but also good for the environment. Our project provides one such solution in the form of Crop health monitoring system. In this project, images are processed and transferred to a deep learning model where the identification of weeds is performed using Yolov5. This model presents an alternative to the old traditional methods of weed detection and help the local communities improve their crop production and revenue