Distribución poblacional de plantaciones de Banano mediante visión por Computadora e Imágenes Aéreas por dron

Abstract

In recent years, Ecuador's banana industry has been affected by various diseases, with Moko being one of the most devastating for farmers. To remain competitive in the global market and ensure the economic stability of thousands of families, it is essential to enhance the management of banana plantations through advanced technologies. In this context, precision agriculture and the use of multispectral drones offer new possibilities to optimize the quantification and monitoring of these plantations, helping to prevent the spread of such diseases. This project proposes the development of an automated system using computer vision and multispectral analysis, employing the Normalized Difference Vegetation Index (NDVI) to improve the accuracy of banana plant population counts and assess the health of the crops. The system consists of three phases: aerial image collection, training of a computer vision model, and analysis of the collected data, all integrated into a web platform for visualization and consultation. The proposed solution will provide banana producers in Ecuador with an advanced tool for efficient monitoring and management of large crop areas, contributing to the sustainable development of the sector. Furthermore, the implementation of this project will reduce operational costs associated with traditional monitoring methods, such as manual plant counting. This will allow farmers to make data-driven, real-time decisions, improving the sustainability of their operations and strengthening the competitiveness of Ecuadorian bananas in international markets. Keywords: Precision agriculture, computer vision, multispectral drones, banana plantations, automated monitoring, NDVI.