Decision-Making Platform for SMART Plantation Agriculture Using Machine Learning and Image Processing

Abstract

Plantation agriculture plays a crucial role in the Sri Lankan economy in terms of both values of production and employment, even though the relative contribution has declined in recent years. Climate variability and volatile commodity prices influence agricultural production and revenue. Production and marketing decisions are frequently based on insufficient knowledge of the specific outcome of that decision. Therefore, most planters are having difficulty with the decision-making process since they are not using high-level technologies and are relying on conventional approaches. As a result, Sri Lankan agriculture and plantation industries are operating at a lower production capacity. The objective of this study is to analyze and propose appropriate solutions to the challenges that the planters face daily based on their environmental characteristics, previous data, and using their mobile phone cameras, planters will be able to make the most precise decisions using high-level technologies. This system presents a software-enabled platform for predicting future yields, forecasting the future market and intermediate buying selling prices, recognizing pests, and providing appropriate treatments, forecasting a fertilizer plan and water delivery according to soil type, and selecting the most suitable crops for cultivation. Aside from that, introducing a platform where planters can sell their crop to local and international customers and planters can communicate with experts and other planters through an agricultural forum. Machine learning, deep learning, and image processing techniques are employed to develop this system.