Lineamentos topográficos em modelos digitais de elevação: uma comparação entre extração manual e computacional usando redes neurais convolucionais com arquitetura U-Net
DOI:
https://doi.org/10.70369/188dww73Palavras-chave:
aprendizado de máquina, interpretação manual de lineamentos, Redes Neurais Convolucionais, U-Net, lineamentos topográficosResumo
Lineamentos topográficos presentes em modelos digitais de elevação (MDE) oferecem boas indicações sobre estruturas e litotipos. Para extrair informação útil desses dados a interpretação é uma etapa crucial, tradicionalmente feita manualmente por um geocientista, em um processo repetitivo, demorado e sem dúvida subjetivo. Recentemente, técnicas de aprendizado de máquina têm sido empregadas para reconhecer padrões e extrair lineamentos de MDEs de maneira semi-automática. Esta técnica tem mostrado potencial de reduzir drasticamente o tempo necessário para a obtenção de lineamentos e reduzir a subjetividade da interpretação em dados adquiridos por sensores remotos. Neste trabalho assumimos a tarefa de extrair lineamentos de dados SRTM (Shutle Radar Topography Mission) na região central da Faixa de dobramentos Ribeira, SE do Brasil, um orógeno Neoproterozóico-cambriano profundamente erodido. Analisamos como a interpretação humana variou ao longo de cinco rodadas de interpretação de lineamentos na mesma área e comparamos os resultados com extrações de lineamentos feitas por uma Rede Neural Convolucional Profunda (RNCP) com arquitetura U-Net treinada para classificar imagens a nível de pixel como lineamento ou não-lineamento. Os resultados mostram que os critérios para extração manual de lineamentos variaram substancialmente ao longo das diferentes rodadas de interpretação. Lineamentos se tornaram menos numerosos e a repetibilidade aumentou em direção às últimas versões. Os hiperparâmetros da RNCP foram definidos em uma série de testes por tentativa e erro, mas entregaram resultados que mimetizam aqueles obtidos manualmente. RNCPs são uma poderosa ferramenta de análise e processamento de dados, mas requerem tempo para construir, anotar, parametrizar, treinar e testar. Neste sentido, se justifica utilizá-las em grandes conjuntos de dados quando não é viável a interpretação manual. Do contrário, é mais prático que um humano desenhe os lineamentos à mão com duas ou mais versões de interpretação para reduzir a subjetividade.
Referências
Akinluyi, F.O., Olorunfemi, M.O., and Bayowa, O.G., 2018, Investigation of the influence of lineaments, lineament intersections and geology on groundwater yield in the basement complex terrain of Ondo State, Southwestern Nigeria: Applied Water Science, v. 8, n. 49, doi: https://doi.org/10.1007/s13201-018-0686-x.
Almeida F.F.M., Hasui Y., Brito Neves, B.B., and Fuck, R.A., 1977, As províncias estruturais do Brasil, in Simpósio de Geologia do Nordeste, 8th, Bol. Esp., 12p.
Almeida, F.F.M., 1967, Origem e evolução da plataforma brasileira: DNPM: Div. Geol. Min: Boletim 241, 36 p.
Almeida, F.F.M., 1976, The system of continental rifts bordering the Santos Basin, Brazil: Anais da Academia Brasileira de Ciências, n. 48, p.15-26.
Almeida, F.F.M.; Hasui, Y., Brito Neves, B.B., and Fuck, R.A., 1981, Brazilian structural provinces: An introduction: Earth-Science Reviews, v. 17, n. 1-2, p. 1-29, doi: https://doi.org/10.1016/0012-8252(81)90003-9.
Almeida, J., Dios, F., Mohriak, W.U., Valeriano, C.D.M., Heilbron, M., Eirado, L.G., and Tomazzoli, E., 2013, Pre-rift tectonic scenario of the eo-cretaceous Gondwana break-up along SE Brazil-SW Africa: insights from tholeiitic mafic dyke swarms: Geological Society, London, Special Publications, v. 369, p. 11-40, doi: https://doi.org/10.1144/SP369.24.
An, Y., Guo, J,, Ye, Q., Childs, C., Walsh, J., and Dong, R., 2021, Deep convolutional neural network for automatic fault recognition from 3D seismic datasets: Computers & Geosciences, v. 153, doi: https://doi.org/10.1016/j.cageo.2021.104776.
Ayalew, L., and Yamagishi, H., 2005, The application of GIS-based logistic regression for landslide susceptibility mapping in the Kakuda-Yahiko Mountains, Central Japan: Geomorphology, v. 65, n. 1-2, p. 15-31, doi: https://doi.org/10.1016/j.geomorph.2004.06.010.
Bashilov, V.I., 1982, Lineaments and petroleum occurrences in the north of the East European Platform: International Geology Review, v. 24, n. 6, p. 715-718, doi: https://doi.org/10.1080/00206818209451013.
Baskin, C., Liss, N., Mendelson, A., and Zheltonozhskii, E, 2017, Streaming Architecture for Large-Scale Quantized Neural Networks on an FPGA-Based Dataflow Platform, in IEEE International Parallel and Distributed Processing Symposium Workshops, Canada, doi: https://doi.org/10.1109/IPDPSW.2018.00032.
Brito Neves, B.B.D., Campos Neto, M.D.C., and Fuck, R.A., 1999, From Rodinia to Western Gondwana: an approach to the Brasiliano-Pan African Cycle and orogenic collage: Episodes, v. 22, n. 3, p. 155-166, doi: https://doi.org/10.18814/epiiugs/1999/v22i3/002.
Cardoso-Fernandes, J., Teodoro, A., Lima, A., Menuge, J.,Brönner, M., and Steiner, R., 2022, Sentinel-1 and ALOS data for lineament extraction: a comparative study, in SPIE Remote Sensing, doi: https://doi.org/10.1117/12.2636117.
Chandra, S., Rao, V.A., Krishnamurthy, N.S., Dutta, S., and Ahmed, S., 2006, Integrated studies for characterization of lineaments used to locate groundwater potential zones in a hard rock region of Karnataka, India: Hydrogeology Journal, n. 14, p. 1042-1051, doi: https://doi.org/10.1007/s10040-006-0097-1.
Clark Jr., S.F., Moore, R.B., Ferguson, E.W., and Picard, M.Z., 1996, Criteria and Methods for Fracture-Trace Analysis of the New Hampshire Bedrock Aquifer: U.S. Geological Survey Open-File Report, n. 96–479, 12 p., doi: https://doi.org/10.3133/ofr96479.
Cogné, N., Cobbold, P., Riccomini, C., and Gallagher, K., 2013, Tectonic setting of the Taubaté Basin (Southeastern Brazil): Insights from regional seismic profiles and outcrop data: Journal of South American Earth Sciences, v. 42, p. 94-204, doi: https://doi.org/10.1016/j.jsames.2012.09.011.
Daynac, J., Bessin, P., Pochat, S., Mourgues, R., and Shumack, S. 2024, A new workflow for mapping dune features (outline, crestline and defects) combining deep learning and skeletonization from DEM-derived data: Geomorphology, v. 463, doi: https://doi.org/10.1016/j.geomorph.2024.109369.
Du, G., Cao, X., Liang, J., Chen, X., and Zhan, Y., 2020, Medical Image Segmentation based on U-Net: A Review: Journal of Imaging Science and Technology, v. 64, p 020508-1 - 020508-12, doi: https://doi.org/10.2352/J.ImagingSci.Technol.2020.64.2.020508.
Enoh, M.A., Okeke, F.I., and Okeke, U.C. 2021, Automatic lineaments mapping and extraction in relationship to natural hydrocarbon seepage in Ugwueme, South-Eastern Nigeria: Geodesy and Cartography, v. 47, n. 1, p. 34-44, doi: https://doi.org/10.3846/gac.2021.12099.
Farahbakhsh, E., Chandra, R., Olierook, H.K.H., Scalzo, R., Clark, C., Reddy, S.M., and Müller, R.D., 2019, Computer vision-based framework for extracting tectonic lineaments from optical remote sensing data: International Journal of Remote Sensing, v. 41, p. 1760–1787, doi: https://doi.org/10.1080/01431161.2019.1674462.
Fletcher, S., and Islam, M.Z., 2018, Comparing sets of patterns with the Jaccard index: Australasian Journal of Information Systems, v. 22, doi: https://doi.org/10.3127/ajis.v22i0.1538.
Freitas, N.C., Almeida, J., Heilbron, M., Cutts, K. and Dussin, I., 2021, The Cabo Frio Thrust: a folded suture zone, Ribeira belt, SE Brazil: Journal of Structural Geology, v. 149, doi: https://doi.org/10.1016/j.jsg.2021.104379.
Froner B., Purves S.J., Lowell J., and Henderson J., 2013, Perception of visual information: the role of colour in seismic interpretation: First Break, v. 31, n. 4, doi: https://doi.org/10.3997/1365-2397.2013010.
Giro J.P., Almeida J., Guedes E., and Bruno H., 2021, Tectonic inheritances in rifts: The meaning of NNE lineaments in the continental rift of SE-Brazil: Journal of South American Earth Sciences, v. 108, doi: https://doi.org/10.1016/j.jsames.2021.103225.
Goldstein, A. and Marshak, S. 1988, Analysis of fracture array geometry (Ch. 12), in Marshak, S. and Mitra, G., eds., Basic Methods of Structural Geology: Prentice Hall, p. 249-268.
Guedes, E., Heilbron, M., Vasconcelos, P.M., Valeriano, C.M., Almeida, J. C. H., Teixeira, W. and Thomaz Filho, A. 2005, K-Ar and 40Ar/39Ar ages of dikes emplaced in the onshore basement of the Santos Basin (Resende Area, SE-Brazil): implications for South Atlantic opening and tertiary reactivation: South American Journal of Earth Sciences, v. 18, n. 3-4, p. 371-382, doi: https://doi.org/10.1016/j.jsames.2004.11.008.
Guedes E., Heilbron M., Valeriano C.M., Almeida J. C. H., and Szatmari P., 2016, Evidence of Gondwana early rifting process recorded by Resende-Ilha Grande Dike Swarm, southern Rio de Janeiro, Brazil: Journal of South American Earth Sciences, v. 67, p. 11-24, doi: https://doi.org/10.1016/j.jsames.2016.01.004.
Guo, Y., Liu, Y., Georgiou, T., and Lew, M.S., 2018, A review of semantic segmentation using deep neural networks: Int Journal Multimed Info Retr, v. 7, p. 87-93, doi: https://doi.org/10.1007/s13735-017-0141z.
Haslebacher, C., Thomas, N., and Bickel, V.T., 2023, LineaMapper: A deep learning-powered tool for mapping linear surface features on Europa: Icarus, v. 410, doi: https://doi.org/10.1016/j.icarus.2023.115722.
Hassanein, A., S., Mohammad, S., Sameer, M., and Ragab, M.E., 2015, A Survey on Hough Transform, Theory, Techniques and Applications: International Journal of Computer Science Issues, v. 12, n. 1, doi: https://doi.org/10.48550/arXiv.1502.02160.
Hasui, Y., Carneiro, C.D.R., and Coimbra, A.M., 1975, The Ribeira folded belt: Revista Brasileira de Geociências, v. 5, n. 4, p. 257-266, doi: https://doi.org/10.25249/0375-7536.1975257266.
Heilbron M., Eirado L.G., and Almeida J.C.H., 2016, Mapa geológico e de recursos minerais do estado do Rio de Janeiro. CPRM: Rio de Janeiro.
Heilbron M., Mohriak, W., Valeriano C.M., Milani E., Almeida J.C.H., and Tupinambá, M., 2000, From collision to extension: the roots of the southeastern continental margin of Brazil, in Talwani and Mohriak., eds., Atlantic Rifts and Continental Margins: America Geophysical Union: Geophysical Monograph Series, v. 115, p. 1-34.
Heilbron, M., Silva, L.G.E., Almeida, J.C.H., et al., 2020, Proterozoic to Ordovician geology and tectonic evolution of Rio de Janeiro State, SE-Brazil: insights on the central Ribeira Orogen from the new 1:400,000 scale geologic map: Brazilian Journal of Geology, v. 50, p. 1-25, doi: https://doi.org/10.1590/2317-4889202020190099.
Heilbron, M., Valeriano, C.M. Valladares, C.S., and Machado, N., 1995, A Orogênese Brasiliana no segmento central da Faixa Ribeira, Brasil: Revista Brasileira de Geociências, v. 25, n. 4, p. 249-266, doi: https://doi.org/10.25249/0375-7536.1996249266.
Hobbs, W.H., 1904, Lineaments of the Atlantic Border region: GSA Bulletin, v. 15, n.1, p. 483-506, doi: https://doi.org/10.1130/GSAB-15-483.
Instituto Nacional de Pesquisas Espaciais, Brasil, 2008, Topodata: Banco de Dados Geomorfométricos do Brasil: http://www.dsr.inpe.br/topodata (accessed Agosto 2025).
Jordan, G., Meijninger, B.M.L., van Hinsbergen, D.J.J., Meulenkamp, J.E., and van Dijk, P.M., 2005, Extraction of morphotectonic features from DEMs: Development and applications for study areas in Hungary and NW Greece: International Journal of Applied Earth Observation and Geoinformation, v. 7, n. 3, p. 163-182, https://doi.org/10.1016/j.jag.2005.03.003.
Liu, C.C., Sousa, M.A., and Gopinath, T.R., 2000, Regional Structural Analysis by Remote Sensing for Mineral Exploration, Paraiba State, Northeast Brazil: Geocarto International, v. 15, n. 1, p. 70-77, doi: https://doi.org/10.1080/10106040008542142.
Lulseged A., and Hiromitsu Y., 2005, The application of GIS-based logistic regression for landslide susceptibility mapping in the Kakuda-Yahiko Mountains, Central Japan: Geomorphology, v. 65, n. 1-2, p. 15-31, doi: https://doi.org/10.1016/j.geomorph.2004.06.010.
Mabee, S.B., Kenneth C.H., Donald U., and Wise, D.U., 1994, A method of collecting and analyzing lineaments for regional‐scale fractured‐bedrock aquifer studies: Groundwater, v. 32, n. 6, p. 884-894, doi: https://doi.org/10.1111/j.1745-6584.1994.tb00928.x.
McCulloch, W.S., and Pitts, W., 1943, A logical calculus of the ideas immanent in nervous activity: Bulletin of Mathematical Biophysics, v. 5, p. 115-133, doi: https://doi.org/10.1007/BF02478259.
Meshkani S.A., Mehrabi B., Yaghubpur A., and Sadeghi M., 2013, Recognition of the regional lineaments of Iran: Using geospatial data and their implications for exploration of metallic ore deposits: Ore Geology Reviews, v. 55, p. 48-63, doi: https://doi.org/10.1016/j.oregeorev.2013.04.007.
Negrão A.P., Mello C.L., Ramos R.R.C., Sanson M.S.R., Louro V.H.A., and Bauli P.G., 2020, Tectonosedimentary evolution of the Resende and Volta Redonda basins (Cenozoic, Central Segment of the Continental Rift of Southeastern Brazil): Journal of South American Earth Sciences, v. 104, doi: https://doi.org/10.1016/j.jsames.2020.102789.
O’Leary, D.W., Friedman J.D., and Pohn, H.A., 1976, Lineament, linear, lineation: some proposed new standards for old terms: GSA Bulletin, v. 87, n. 10, p. 1463-1469, doi: https://doi.org/10.1130/0016-7606(1976)87%3C1463:LLLSPN%3E2.0.CO;2.
Peacock, D.C., Sanderson, D.J., Bastesen, E., Rotevatn, A. and Storstein, T.H., 2019, Causes of bias and uncertainty in fracture network analysis: Norwegian Journal of Geology, v. 99, n. 1, p. 113-128, doi: https://doi.org/10.17850/NJG99-1-06.
Ramli, M.F., Yusof, N., Yusoff, M.K., Juahir, H., and Shafri H.Z.M., 2010, Lineament mapping and its application in landslide hazard assessment: a review: Bull Eng Geol Environ, v. 69, p. 215-233, doi: https://doi.org/10.1007/s10064-009-0255-5.
Riccomini, C., Sant’Anna, L.G., and Ferrari, A.L., 2004, Evolução geológica do rift continental do sudeste do Brasil, in Mantesso-Neto, V., Bartorelli, A., Dal Ré Carneiro, C., Brito Neves, orgs., B.B., eds., Geologia do continente sul-americano – evolução da obra de Fernando Flávio Marques de Almeida. São Paulo, Brasil: Beca.
Ronneberger, O., Fischer, P., and Brox, T., 2015, U-Net: Convolutional Networks for Biomedical Image Segmentation, in Navab, N., Hornegger, J., Wells, W., Frangi, A., eds., Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015: Lecture Notes in Computer Science, v. 9351, doi: https://doi.org/10.1007/978-3-319-24574-4_28.
Saha, S.K., 2022, Remote Sensing and Geographic Information System Applications in Hydrocarbon Exploration: a Review: Journal Indian Soc Remote Sens, v. 50, p. 1457-1475, doi: https://doi.org/10.1007/s12524-022-01540-9.
Sander, P., 2007, Lineaments in groundwater exploration: a review of applications and limitations: Hydrogeology Journal, v. 15, p. 71-74, doi: https://doi.org/10.1007/s10040-006-0138-9.
Schmitt R.S., Trouw R.A.J., Schmus W.R.V.S., and Pimentel M.M., 2004, Late amalgamation in the central part of West Gondwana: new geochronological data and the characterization of a Cambrian collisional orogeny in the Ribeira Belt (SE Brazil): Precambrian Research, v. 133, n. 1-2, p. 29-61, doi: https://doi.org/10.1016/j.precamres.2004.03.010.
Siddique, N., Sidike, P., Elkin, C., and Devabhaktuni, V. 2021, U-Net and Its Variants for Medical Image Segmentation: A Review of Theory and Applications: IEEE Access, v. 9, p. 82031-82057, doi: https://doi.org/10.1109/ACCESS.2021.3086020.
Trouw, R., Heilbron, M., Ribeiro, A., Paciullo, F., Valeriano, C.; Almeida, J.C.H., Tupinambá, M., and Andreis, R., 2000, The Central Segment of The Ribeira Belt, in Cordani, U.G., Milani, E.J., Thomás Filho, A. and Campos, D.A., eds., Tectonic Evolution of South America, International Geological Congress, 31ed., Rio de Janeiro: Academia Brasileira de Ciências, p. 287-310.
Yusof, N., Ramli, M.F., Pirasteh S., and Shafri, H.Z.M., 2011, Landslides and lineament mapping along the Simpang Pulai to Kg Raja highway, Malaysia: International Journal of Remote Sensing, v. 32, n. 14, p. 4089-4105, dois: https://doi.org/10.1080/01431161.2010.484434.
Zhang, Y., Gang W., Mingchao L., and Shuai H., 2018, Automated Classification Analysis of Geological Structures Based on Images Data and Deep Learning Model: Applied Sciences, v. 8, n. 12, p. 2493, doi: https://doi.org/10.3390/app8122493.
Downloads
Publicado
Declaração de Disponibilidade de Dados
Os dados utilizados são públicos, e estão disponíveis através do link:
Edição
Seção
Licença
Copyright (c) 2025 Pedro Rossi, Julio Cesar Horta de Almeida, Antônio de Pádua Pires, Mario Neto Araujo
Este trabalho está licenciado sob uma licença Creative Commons Attribution 4.0 International License.
Esta licença permite que os reutilizadores distribuam, remixem, adaptem e criem obras derivadas do material em qualquer meio ou formato, desde que seja dada a devida atribuição ao criador. A licença permite o uso comercial.
