Satellite images and deep learning for the prediction of socioeconomic indicators in the Vale do Ribeira

• Jeaneth Machicao USP
• Iago Fava da Costa USP
• Enrico Triñanes USP
• Pedro Corrêa USP

DOI: https://doi.org/10.5753/bresci.2022.222511

Resumo

Due to low awareness and low investment in data collection and processing in developing countries, industry has struggled to plan and collaborate with public policy on data-driven socioeconomic decisions. This paper presents a deep learning approach to estimate socioeconomic indicators using satellite imagery. For this purpose, a region in southeastern Brazil called Vale do Ribeira was selected as a study case due to its data availability and environmental relevance. We used publicly available data to analyze three socioeconomic indicators: Longevity, Income and Literacy, the main representatives of the Human Development Index. Our preliminary results show a relevant correlation between satellite imagery and the income indicator, although they are not yet conclusive for longevity and literacy.

Palavras-chave: deep learning, remote sensing imagery, poverty estimation

Referências

Abreu, M., Oliveira, J., Andrade, V., & Meira, A. (2011). Methodological proposal for spatial calculation and analysis of the intra-urban HDI of Viçosa, Brazil. Revista Brasileira de Estudos de População. 28. 169-186.

Ayush, K., Uzkent, B., Tanmay, K., Burke, M., Lobell, D., & Ermon, S. (2021). Efficient Poverty Mapping from High Resolution Remote Sensing Images. Proceedings of the AAAI Conference on Artificial Intelligence, 35(1), 12-20.

Blumenstock, J. E. (2016). Fighting poverty with data. Science, 353(6301), 753-754.

Bueno, G.W., Leonardo, A.F.G., Machado, L.P., Brande, M.R., Godoy, E.M., & David, F.S. (2020). Indicadores de sustentabilidade socioambiental de pisciculturas familiares em Área de Mata Atlântica, no Vale do Ribeira - SP. Arquivo Brasileiro de Medicina Veterinária e Zootecnia, 72(3), 901-910.

Burke, M., Driscoll, A., Lobell, D. B., & Ermon, S. (2021). Using satellite imagery to understand and promote sustainable development. Science, 371(6535).

Deng, J., Dong, W., Socher, R., Li, L-J., Kai, L., & Li F-F. (2009). ImageNet: A largescale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, 248-255.

Gorelick, N., Hancher, M., Dixon, M., Ilyushchenko, S., Thau, D., & Moore, R. (2017). Google Earth Engine: Planetary-scale geospatial analysis for everyone. Remote Sensing of Environment, 202, 18-27.

Jean, N., Burke, M., Xie, M., Davis, W. M., Lobell, D. B., & Ermon, S. (2016) Combining satellite imagery and machine learning to predict poverty. Science, 353(6301), 790–794.

Machicao, J., Specht, A., Vellenich, D., Meneguzzi, L., David, R., Stall, S., Ferraz, K., Mabile, L., O'Brien, M. & Corrêa, P. (2022). A Deep-Learning Method for the Prediction of Socio-Economic Indicators from Street-View Imagery Using a Case Study from Brazil. Data Science Journal, 21(1), p.6.

National Geophysical Data Center. (2010). Version 4 DMSP-OLS Nighttime Lights Time Series.

Planet Team (2017). Planet Application Program Interface: In Space for Life on Earth. San Francisco, CA. https://api.planet.com

PNUD. Índice de Desenvolvimento Humano Municipal - IDHM: Metodologia. (2012). https://atlasbrasil.org.br/acervo/atlas

Simonyan, K.; Zisserman, A. (2014). Very Deep Convolutional Networks for Large-Scale Image Recognition. https://arxiv.org/abs/1409.1556.

Suel, E., Polak, J.W., Bennett, J.E., & Ezzati, M. (2019). Measuring social, environmental and health inequalities using deep learning and street imagery. Scientific Reports, 9(1), 6229.

Yeh, C., Perez, A., Driscoll, A., Azzari, G., Tang, Z., Lobell, D., Ermon, S., & Burke, M. (2020). Using publicly available satellite imagery and deep learning to understand economic well-being in Africa. Nature Communications, 11.