Abstract

Smart agriculture has surfaced as a revolutionary method to enhance crop productivity and sustainability. One critical aspect of this paradigm is the integration of Edge- Artificial Intelligence (Edge-AI) for efficient plant disease identification and prevention. This study investigates the latest developments in employing Edge-AI techniques for the control of plant leaf diseases in smart agriculture. The article commences by providing a synopsis of the challenges posed by plant diseases in agriculture and their potential implications for global food security. It underscores the necessity for proactive identification of diseases and prevention strategies to mitigate these challenges. The survey then delves into the integration of Edge-AI technologies at the edge of the agricultural ecosystem, enabling real-time processing and decision-making. The analysis encompasses a thorough examination of diverse Edge-AI models and algorithms utilized for identifying plant leaf diseases, highlighting both their strengths and limitations. Principal techniques, such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and ensemble learning methods are emphasized in the context of their effectiveness in identifying diverse types of plant diseases. The survey highlights the importance of edge computing in enabling localized processing and decision-making, reducing latency and dependency on centralized systems. It explores the integration of sensor networks, Internet of Things (IoT) devices, and Edge-AI in creating a robust and responsive smart agriculture ecosystem. The paper also addresses challenges and open research issues in the field, including data privacy concerns, model robustness, and the need for standardized datasets. Additionally, it discusses emerging trends such as federated learning and blockchain for secure and collaborative disease management. This study offers an extensive overview of the cutting-edge Edge-AI technologies for identifying and preventing plant leaf diseases in smart agriculture. By comprehending the present landscape and anticipating future directions, researchers and practitioners can make informed decisions to create scalable and efficient solutions for sustainable and resilient agricultural systems.