Classification of wetlands in multispectral satellite images of the Arequipa region using artificial intelligence algorithms

Cornejo Paredes A.S.; Paja Medina A.I.; Rosas Paredes K.; Zuñiga Carnero M.; Zuñiga Portilla E.; Sulla Torres J.A.; Peyre G. · CISCI - Conferencia Iberoamericana en Sistemas, Cibernetica e Informatica (2024) · DOI: 10.54808/cisci2024.01.315

Resumen

At a global level, given their ecological importance, there is a need to map the presence of wetlands due to their critical role in the conservation of fragile ecosystems. However, current assessment faces significant challenges due to costly and subjective methods, which complicate obtaining accurate data. In various regions, wetlands, ecosystems rich in biodiversity, are threatened by climate change and environmental degradation. The lack of monitoring and technological knowledge limits the ability to make informed decisions and thoroughly study these vital areas. This research aims to classify wetlands in multispectral satellite images of the Arequipa region by applying supervised learning algorithms to contribute to scientific knowledge with a better understanding of the distribution of wetlands. The results showed that the CNN algorithm is the most accurate, followed by RF and KNN for classifying wetlands in the Arequipa region. © CISCI - Simposium Iberoamericano en Educacion, Cibernetica e Informatica 2024.

Instituciones: Universidad Católica de Santa María