AI-Driven Detection of Plastic Materials in Agricultural Environments

• Grant number: 24/08841-9
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: July 01, 2024
• End date: June 30, 2025
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Agreement: BRASKEM
• Principal Investigator: Rubens Augusto Camargo Lamparelli
• Grantee: Igor Engelmann Batista
• Host Institution: Núcleo Interdisciplinar de Planejamento Energético (NIPE). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Company: Universidade Estadual de Campinas (UNICAMP). Núcleo Interdisciplinar de Planejamento Energético (NIPE)
• Associated research grant: 21/05251-8 - Proposal for the creation of the Plasticulture Engineering Centre, AP.PCPE

Abstract

The use of plastic in agriculture has boosted the productivity of crops or the quality of products. On the other hand, an informational gap exists about its distribution in Brazil. A solution for this situation can reside in applying AI-driven Computer Vision technologies to analyze satellite imagery of crops, aiming to map plastic structures in agricultural areas. We aim to develop an enhanced computer vision model utilizing deep learning AI-driven solutions tailored for identifying plastic structures in agricultural settings with better metrics than the solutions found in the literature. We also seek to enhance the integration of these image recognition systems with Geographic Information Systems to improve data quality for Circular Economy route planning and sustainable resource management. To achieve this goal the selected network will be trained with labeled images of melon-growing areas that use plastic mulching. We selected the region of interest in Rio Grande do Norte and Ceará states, Brazil due to the significant plasticulture area and the computational challenges imposed by the applied management and climatic conditions. The expected result is an ANN model with consistent efficiency amidst varying luminosity levels and environmental conditions, making it applicable across diverse locations and contexts.