Intelligent weed management using aerial image processing and precision herbicide spraying: An overview

Modern agriculture is increasingly adopting intelligent technologies to enhance productivity while minimizing production costs and reducing adverse environmental impacts. A prime example of this synergy is the use of image processing to identify weeds, enabling targeted herbicide spraying with autonomous devices such as robots and drones. This approach not only reduces production costs but also ensures sustainable farming while minimizing negative environmental impacts. Designing an intelligent weed management system requires a multidisciplinary approach, combining agriculture, big data processing, machine learning, computer science, robotics, and plant science. Currently, independent studies have focused on some of these aspects, but few have taken a holistic approach to address the issue. This paper highlights the approach taken in developing innovative and ecologically sustainable weed management systems for agriculture. It also presents a comprehensive overview of a weed management system that integrates coordinated weed detection and spraying, detailing its unique components. The paper reviews and contrasts various image analysis techniques used in weed detection, particularly those employing artificial intelligence and imagery captured by unmanned aerial vehicles (UAVs). Furthermore, the paper highlights recent advancements in image processing platforms, such as the shift towards local and edge computing, and the growing need for near-real-time processing in agricultural applications. It also explores the development of commercial weed-spraying drones and discusses various aspects of an autonomous weed control system, including design, navigation, and spraying mechanisms for targeted application. Finally, the paper identifies key research needs for developing an AI-based, targeted herbicide spraying system that could significantly contribute to sustainable, economically viable, and efficient agricultural practices. • Novel automated spraying model advances precision agriculture research. • Review: UAV image processing & ML for accurate weed identification. • Challenges & future of automated herbicide systems in precision farming.