Web System Based on Machine Learning to Evidence Reforestation Efforts Through the Processing of Satellite Images of the Amazon

Castro R.A.H.; Hernández O.A.R.; Alarcon J.S.D. · International Conference on Artificial Intelligence, Computer, Data Sciences, and Applications, ACDSA 2026 (2026) · DOI: 10.1109/acdsa67686.2026.11467842

Resumen

The Peruvian Amazon plays a critical role in global climate regulation and carbon sequestration, but it faces accelerating deforestation driven by human activities. The evaluation of reforestation projects is challenging due to the size of the region and the limitations of conventional monitoring. This study presents a web-based approach that combines satellite multispectral imagery from Sentinel-2 with structural information from the Global Ecosystem Dynamics Investigation lidar mission to provide an objective assessment of ecological restoration efforts. The methodology comprises four phases: benchmarking of candidate algorithms and datasets; selection of key variables, including vegetation indices and spectral bands, along with structural attributes of the forest; implementation of the system; and validation of the model with standard quantitative metrics. In the retained data, the system achieves a coefficient of determination of 0.84 for the estimation of forest biomass, which indicates a high predictive utility. These results suggest an effective and scalable tool to support evidence-based decision-making for reforestation and conservation monitoring in the region.

Instituciones: Universidad Peruana de Ciencias Aplicadas