Integrated Approach for Disease Diagnosis in Fruit Crops: Economic Analysis, Technological Review, and Development of an Artificial Intelligence-Based System
Fruit Farming, Artificial Intelligence, Disease Diagnosis, Value, Convolutional Neural Networks.
Brazilian fruit farming stands out on the global stage, positioning the country as the third-largest producer worldwide. This study addresses the economic dimension of the sector in Mato Grosso, which, through the analysis of the Gross Production Value (GPV) between 2019 and 2023, identified bananas, watermelons, and pineapples as the most relevant crops, jointly ac-counting for more than 80% of the state’s fruit farming GPV. In view of the phytosanitary chal-lenges that affect productivity, a systematic review was conducted based on the PRISMA 2020 methodology, analyzing 48 publications (2020–2024) on the use of Artificial Intelligence (AI) for disease diagnosis in these three crops. The results revealed a predominance of research fo-cused on bananas and the superiority of Convolutional Neural Networks (CNNs), which achi-eved accuracy rates above 95% in most cases. The analysis highlighted significant research gaps for watermelon and, especially, pineapple crops. To address this gap, the final phase of the work proposes the prototyping and evaluation of a CNN-based system for the automatic iden-tification of leaf diseases in pineapples, aiming to provide an accessible and practical decision-support tool for farmers.