Deep Learning for Weed Detection: Exploring YOLO V8 Algorithm's Performance in Agricultural Environments
Pushpendra Kumar, Upendra Misra
Abstract
The YOLO V8 algorithm is a state-of-the-art deep one-stage object detection algorithm. This study assesses its effectiveness in weed detection in agricultural environments. The algorithm's performance in identifying and localizing weed species within crop fields was evaluated using a diverse dataset of crops and weed species. The YOLO V8 algorithm achieved an accuracy of 86% in weed identification in the real-time agricultural environment, with minimal false positives. As it is a single-shot object detection algorithm it exhibits excellent processing speeds, making it suitable for practical field applications. The findings demonstrate the potential of deep learning techniques for robust and real-time weed detection in agricultural settings. The weed management practices can be benefited from the help of machine vision and deep learning techniques and contribute towards sustainable development.