Machine Learning (ML) Algorithms on IoT and Drone Data for Smart Farming

Morocco’s GDP is 15% agriculture. Agricultural diseases can devastate national economies. Detecting diseases early reduces their impact and saves more crops. The global population will increase to 9.6 billion people by 2050, thanks in large part to 200,000 new people every day. We can only feed the world’s expanding population with increased food yields. Agricultural modernization is necessary. Manually identifying diseases is time-consuming, error-prone, and challenging. Automation saves time and labor. This study describes recent machine learning, deep learning, image processing, Internet of Things (IoT), and hyperspectral image analysis research in agricultural disease identification. The “Internet of Things” connects common things, machinery, cars, and other electrical gadgets to the web to share data (IoT). The Internet of Things is used to link gadgets and collect data. Farmers can benefit greatly from the IoT. The project aims to develop an IoT- powered smart farm. Farmers may be able to increase crop yields with drone technology by overcoming various hurdles. Drone technology has commercial and surveying applications. This research helped build a drone-based system for detecting and analyzing leaf diseases in agriculture. Globally, “drone” refers to unmanned aerial vehicles (UAVs). Identifying plant diseases accurately is vital to reducing agricultural productivity and quantity losses. Computers with a radio frequency transmitter and receiver system allow the remote operation of “drones.” Diseases that leave a pattern on the leaves make studying them easier. Agricultural drones are used for fertilizer spraying, seed sowing, crop monitoring, and mapping. Automatic leaf photography is feasible with a drone and Raspberry Pi camera. Photos can identify disease indicators. Precision agriculture (PA) uses for drones include crop monitoring and pesticide spraying. Building drones, enhancing sensors, and spraying are explained. AI and deep learning are also possible remote agriculture monitoring systems. Various crop disease diagnosis methodologies were also compared. This research also solves several problems. After problems are identified, solutions are given. This finding should help scientists detect agricultural illnesses in the future.