Técnicas de Machine Learning Aplicadas à Classificação de Pragas na Cultura de Soja

Abstract

The evolution of agriculture over the centuries has culminated in Agriculture 4.0, characterized by the integration of connectivity and automation into the sector. In this context, Unmanned Aerial Vehicles (UAVs), commonly known as drones, emerge as valuable tools for Agriculture 4.0. Drones enable efficient property mapping, remote monitoring of crops, and early detection of issues. Leveraging Machine Learning techniques, drones can accurately identify pests and diseases. This innovative approach optimizes production and reduces costs. The objective of this study is to develop an Artificial Neural Network (ANN) based identification model capable of detecting specific pests, such as caterpillars and the green-yellow beetle (Diabrotica speciosa), in soybean plants. The model aims to assist decision-making regarding the application of agricultural pesticides by identifying affected areas and pinpointing the insect responsible for the damage.