Clasificación de zonas agrícolas en Colombia por medio de imágenes satelitales con redes neuronales profundas

dc.contributor.advisorGonzález Osorio, Fabio Augustospa
dc.contributor.advisorRamos Pollán, Raúlspa
dc.contributor.authorÁlvarez Montoya, Sebastián Felipespa
dc.contributor.researchgroupMindlabspa
dc.coverage.countryColombiaspa
dc.coverage.tgnhttp://vocab.getty.edu/page/tgn/1000050
dc.date.accessioned2024-04-02T00:30:16Z
dc.date.available2024-04-02T00:30:16Z
dc.date.issued2023
dc.descriptionilustraciones, diagramas, fotografías, mapas, planosspa
dc.description.abstractLas imágenes satelitales son una fuente valiosa de información sobre la tierra, que nos permiten analizar su superficie y las estructuras creadas por el ser humano, como la cobertura del suelo, la vegetación, la topografía y las áreas urbanas. En las últimas décadas, se han producido avances significativos para mejorar la calidad de estas imágenes, incluyendo el uso de imágenes multiespectrales de alta resolución que brindan una descripción más precisa de los objetos y su entorno. Además, se han desarrollado modelos de aprendizaje profundo utilizando estas imágenes para la clasificación y segmentación de objetos; principalmente en ámbitos urbanos y climáticos, con pocos modelos enfocados en la agricultura y los cultivos. Sin embargo, dado que Colombia es un país con una vasta extensión de tierra dedicada a la agricultura, es importante desarrollar modelos de aprendizaje profundo para clasificar y predecir la distribución de estas áreas, lo que brinda información valiosa tanto al gobierno como a los agricultores. En este estudio se utilizaron imágenes de los satélites Sentinel 2 tomadas en el año 2020, que fueron preprocesadas y georreferenciadas. Luego se determinó la cantidad de área porcentual de zonas agrícolas en cada imagen, que es la variable que permite la etiquetación de las mismas, como Frontera agrícola o No en la tarea de clasificación. Se utilizaron redes neuronales convolucionales profundas, incluyendo MobileNet, ResNet50, Inception v3 y VGG 19, con una entrada de resolución de imagen de 100 x 100. De igual manera, se utilizaron modelos con arquitecturas más simples para hacer una comparación adicional entre estos tipos de modelos; los cuales se dividieron como modelos shallow convolutional y modelos basados en Quantum Kernel Mixtures. Donde se observan mejores resultados utilizando estas arquitecturas más simples para esta tarea de clasificación con este tipo de imágenes. En resumen, este estudio demuestra cómo el uso de modelos de aprendizaje profundo junto con imágenes satelitales de alta resolución puede proporcionar información valiosa para la agricultura, permitiendo una mejor comprensión y planificación de las áreas de cultivo en Colombia. (Texto tomado de la fuente).spa
dc.description.abstractSatellite images constitute a valuable source of information about the Earth, enabling the analysis of its surface and human-created structures, such as land cover, vegetation, topography, and urban areas. Significant advancements have been made in recent decades to enhance the quality of these images, including the utilization of high-resolution multispectral images that provide a more precise description of objects and their surroundings. Additionally, deep learning models have been developed using these images for object classification and segmentation, primarily in urban and climatic contexts, with limited focus on agriculture and crops. Given that Colombia encompasses vast agricultural lands, it is crucial to develop deep learning models for classifying and predicting the distribution of these areas, offering valuable insights to both the government and farmers. This study utilized images from Sentinel 2 satellites captured in the year 2020, which underwent preprocessing and georeferencing. The percentage of agricultural area in each image was then determined, serving as the variable for labeling them as either agricultural land or No in the classification task. Deep convolutional neural networks, including MobileNet, ResNet50, Inception v3, and VGG 19, were employed with an input image resolution of 100 x 100. Similarly, models with simpler architectures were used for additional comparison, categorized as shallow convolutional models and models based on Quantum Kernel Mixtures. Interestingly, superior results were observed using these simpler architectures for this classification task with high-resolution satellite images. In summary, this study demonstrates how the combination of deep learning models and high-resolution satellite images can provide valuable information for agriculture, facilitating a better understanding and planning of cultivation areas in Colombia.eng
dc.description.degreelevelMaestríaspa
dc.description.degreenameMagíster en Ingeniería - Ingeniería de Sistemas y Computaciónspa
dc.description.researchareaComputación aplicada - Sistemas inteligentesspa
dc.format.extentx, 71 páginasspa
dc.format.mimetypeapplication/pdfspa
dc.identifier.instnameUniversidad Nacional de Colombiaspa
dc.identifier.reponameRepositorio Institucional Universidad Nacional de Colombiaspa
dc.identifier.repourlhttps://repositorio.unal.edu.co/spa
dc.identifier.urihttps://repositorio.unal.edu.co/handle/unal/85836
dc.language.isospaspa
dc.publisherUniversidad Nacional de Colombiaspa
dc.publisher.branchUniversidad Nacional de Colombia - Sede Bogotáspa
dc.publisher.facultyFacultad de Ingenieríaspa
dc.publisher.placeBogotá, Colombiaspa
dc.publisher.programBogotá - Ingeniería - Maestría en Ingeniería - Ingeniería de Sistemas y Computaciónspa
dc.relation.indexedAgrosaviaspa
dc.relation.indexedAgrovocspa
dc.relation.referencesDe Agricultura, Ministerio (Ed.): Metodología para la identificación general de la frontera agrícola en Colombia. Unidad de Planificación Rural Agropecuaria, 2018spa
dc.relation.referencesChatterjee, A. ; Sha, J. ; Mukherjee, J. ; Aikat, S. ; Misra, A.: Unsupervised land cover classification of hybrid and dual-polarized images using deep convolutional neural network. En: IEEE Geoscience and Remote Sensing Letters 18 (6) (2021), p. 969–973spa
dc.relation.referencesChen, Z. ; Wang, Y. ; Han, W. ; Feng, R. ; Chen, J.: An improved pretraining strategy-based scene classification with deep learning. En: IEEE Geoscience and Remote Sensing Letters 15 (5) (2020), p. 844–848spa
dc.relation.referencesCheng, G. ; Xie, X. ; Han, J. ; Guo, L. ; Xia, G.: Remote sensing image scene classification meets deep learning: challenges, methods, benchmarks, and opportunities. En: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 13 (2020), p. 3735–3756spa
dc.relation.referencesDai, X. ; Wu, X. ; Wang, B. ; Zhang, L.: Semisupervised scene classification for remote sensing images: a method based on convolutional neural networks and ensemble learning. En: IEEE Geoscience and Remote Sensing Letters 16 (6) (2019), p. 869–873spa
dc.relation.referencesGao, Y. ; Li, Q.: A segmented particle swarm optimization convolutional neural network for land cover and land use classification of remote sensing images. En: Remote Sensing Letters 10 (12) (2019), p. 1182–1191spa
dc.relation.referencesGonzález, F. ; Ramos-Pollán, R. ; Gallego-Mejia, J.: Kernel density matrices for probabilistic deep learning. En: arXiv: 2305.18204v2 (2023)spa
dc.relation.referencesHan, Y. ; Liu, Y. ; Hong, Z. ; Zhang, Y. ; Yang, S. ; Wang, J.: Sea ice image classification based on heterogeneous data fusion and deep learning. En: Remote Sensing 13 (4) (2021), p. 1–20spa
dc.relation.referencesHaykin, S.: Neural Networks and Learning Machines, 3rd edition. Prentice Hall, 2008spa
dc.relation.referencesHe, K. ; Zhang, X. ; Ren, S. ; Sun, J.: Deep residual learning for image recognition. En: arXiv: 1512.03385v1 (2015)spa
dc.relation.referencesHelber, P. ; Bischke, B. ; Dengel, A. ; Borth, D.: Eurosat: A novel dataset and deep learning benchmark for land use and land cover classification. En: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 12 (7) (2019), p. 2217–2226spa
dc.relation.referencesHong, D. ; Gao, L. ; Yokoya, N. ; Yao, J. ; Chanussot, J. ; Du, Q. ; Zhang, B.: More diverse means better: multimodal deep learning meets remote-sensing imagery classification. En: IEEE Transactions on Geoscience and Remote Sensing 59 (5) (2021), p. 4340–4354spa
dc.relation.referencesHoward, A. ; Zhu, M. ; Chen, B. ; Kalenichenko, D. ; Wang, W. ; Weyand, T. ; Andreetto, M. ; Adam, H.: MobileNets: Efficient convolutional neural networks for mobile version applications. En: arXiv: 1704.04861v1 (2017)spa
dc.relation.referencesIllarionova, S. ; Trekin, A. ; Ignatiev, V. ; Oseledets, I.: Neural-based hierarchical approach for detailed dominant forest species classification by multispectral satellite imagery. En: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 14 (2021), p. 1810–1820spa
dc.relation.referencesJamali, A. ; Mahdianpari, M. ; Brisco, B. ; Granger, J. ; Mohammadimanesh, F. ; Salehi, B.: Comparing solo versus ensemble convolutional neural networks for wetland classification using multi-spectral satellite imagery. En: Remote Sensing 13 (11) (2021)spa
dc.relation.referencesJozdani, S. ; Johnson, B. ; Chen, D.: Comparing deep neural networks, ensemble classifiers, and support vector machine algorithms for object-based urban landuse/land cover classification. En: Remote Sensing 11 (14) (2019)spa
dc.relation.referencesKussul, N. ; Lavreniuk, M. ; Skakun, S. ; Shelestov, A.: Deep learning classification of land cover and crop types using remote sensing data. En: IEEE Geoscience and Remote Sensing Letters 14 (5) (2017), p. 778–782spa
dc.relation.referencesDe Lima, R. ; Marfurt, K.: Convolutional neural network for remote-sensing scene classification: transfer learning analysis. En: Remote Sensing 12 (1) (2020)spa
dc.relation.referencesLiu, Q. ; Hang, R. ; Song, H. ; Li, Z.: Learning multiscale deep features for highresolution satellite image scene classification. En: IEEE Transactions on Geoscience and Remote Sensing 56 (1) (2018), p. 117–126spa
dc.relation.referencesLiu, X. ; He, C. ; Zhang, Q. ; Liao, M.: Statistical convolutional neural network for land-cover classification from SAR images. En: IEEE Geoscience and Remote Sensing Letters 17 (9) (2020), p. 1548–1552spa
dc.relation.referencesLiu, X. ; Zhou, Y. ; Zhao, J. ; Yao, R. ; Liu, B. ; Zheng, Y.: Siamese convolutional neural networks for remote sensing scene classification. En: IEEE Geoscience and Remote Sensing Letters 16 (8) (2019), p. 1200–1204spa
dc.relation.referencesLiu, Y. ; Zhong, Y. ; Qin, Q.: Scene classification based on multiscale convolutional neural network. En: IEEE Transactions on Geoscience and Remote Sensing 56 (12) (2018), p. 7109–7121spa
dc.relation.referencesMaggiori, E. ; Tarabalka, Y. ; Charpiat, G. ; Alliez, P.: Convolutional neural networks for large-scale remote-sensing image classification. En: IEEE Transactions on Geoscience and Remote Sensing 55 (2) (2017), p. 645–657spa
dc.relation.referencesPeng, C. ; Li, Y. ; Jiao, L. ; Shang, R.: Efficient convolutional neural architecture search for remote sensing image scene classification. En: IEEE Transactions on Geoscience and Remote Sensing 59 (7) (2021), p. 6092–6105spa
dc.relation.referencesQin, S. ; Guo, X. ; Sun, J. ; Qiao, S. ; Zhang, L. ; Yao, J. ; Cheng, Q. ; Zhang, Y.: Landslide detection from open satellite imagery using distant domain transfer learning. En: Remote Sensing 13 (11) (2021)spa
dc.relation.referencesRosso, M. D. ; Sebastianelli, A. ; Spiller, D. ; Mathieu, P. ; Ullo, S.: Onboard volcanic eruption detection through CNNs and satellite multispectral imagery. En: Remote Sensing 13 (17) (2021)spa
dc.relation.referencesScott, G. ; England, M. ; Starms, W. ; Marcum, R. ; Davis, C.: Training deep convolutional neural networks for land-cover classification of high-resolution imagery. En: IEEE Geoscience and Remote Sensing Letters 14 (4) (2017), p. 549–553spa
dc.relation.referencesShendryk, Y. ; Rist, Y. ; Ticehurst, C. ; Thorburn, P.: Deep learning for multimodal classification of cloud, shadow and land cover scenes in PlanetScope and Sentinel-2 imagery. En: IEEE Geoscience and Remote Sensing Letters 157 (2019), p. 124–136spa
dc.relation.referencesShi, C. ; Lv, Z ; Shen, H. ; Fang, L. ; You, Z.: Improved metric learning with the CNN for very-high-resolution remote sensing image classification. En: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 14 (2021), p. 631–644spa
dc.relation.referencesSiesto, G. ; Fern´andez-Sellers, M. ; Lozano-Tello, A.: Crop classification of satellite imagery using synthetic multitemporal and multispectral images in convolution neural networks. En: Remote Sensing 13 (17) (2021)spa
dc.relation.referencesSimonyan, K. ; Zisserman, A.: Very deep convolutional networks for large-scale image recognition. En: arXiv: 1409.1556v6 (2015)spa
dc.relation.referencesSzegedy, C. ; Vanhoucke, V. ; Ioffe, S. ; Shlens, J.: Rethinking the inception architecture for computer vision. En: arXiv: 1512.00567v3 (2015)spa
dc.relation.referencesTang, X. ; Ma, Q. ; Zhang, X. ; Liu, F. ; Ma, J. ; Jiao, L.: Attention consistent network for remote sensing scene classification. En: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 14 (2021), p. 2030–2045spa
dc.relation.referencesTao, C. ; Lu, W. ; Qi, J. ; Wang, H.: Spatial information considered network for scene classification. En: IEEE Geoscience and Remote Sensing Letters 18 (6) (2021), p. 948–988spa
dc.relation.referencesTu, B. ; Kuang, W. ; He, W. ; Zhang, G. ; Peng, Y.: Robust learning of mislabeled training samples for remote sensing image scene. En: IEEE Geoscience and Remote Sensing Letters 18 (2) (2021), p. 241–245spa
dc.relation.referencesWang, G. ; Fan, B. ; Xiang, S. ; Pan, C.: Aggregating rich hierarchical features for scene classification in remote sensing imagery. En: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 10 (9) (2017), p. 4104–4115spa
dc.relation.referencesWu, Z. ; Hou, B. ; Jiao, L.: Multiscale CNN with autoencoder regularization joint contextual attention network for SAR image classification. En: IEEE Transactions on Geoscience and Remote Sensing 59 (2) (2021), p. 1200–1213spa
dc.relation.referencesXu, X. ; Li, W. ; Ran, Q. ; Du, Q. ; Gao, L. ; Zhang, B.: Multisource remote sensing data classification based on convolutional neural network. En: IEEE Transactions on Geoscience and Remote Sensing 56 (2) (2018), p. 937–949spa
dc.relation.referencesYang, N. ; Tang, H. ; Sun, H. ; Yang, X.: Dropband: a simple and effective method for promoting the scene classification accuracy of convolutional neural networks for VHR remote sensing imagery. En: IEEE Geoscience and Remote Sensing Letters 15 (2) (2018), p. 257–261spa
dc.relation.referencesYoo, C. ; Han, D. ; Im, J. ; Bechtel, B.: Comparison between convolutional neural networks and random forest for local climate zone classification in mega urban areas using landsat images. En: ISPRS Journal of Photogrammetry and Remote Sensing 157 (2019), p. 155–170spa
dc.relation.referencesYu, X. ; Wu, X. ; Luo, C. ; Ren, P.: Deep learning in remote sensing scene classification: a data augmentation enhanced convolutional neural network framework. En: GIScience and Remote Sensing 54 (5) (2017), p. 741–758spa
dc.relation.referencesZhang, C. ; Sargent, I. ; Pan, X. ; Li, H. ; Gardiner, A. ; Hare, J. ; Atkinson, P.: An object-based convolutional neural network (OCNN) for urban land use classification. En: Remote Sensing of Environment 216 (2018), p. 57–70spa
dc.relation.referencesZhang, C. ; Wang, X. ; Ma, L. ; Lu, X.: Tropical cyclone intensity classification and estimation using infrared satellite images with deep learning. En: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 14 (2021), p. 2070–2086spa
dc.relation.referencesZhang, Y. ; Zheng, X. ; Yuan, Y. ; Lu, X.: Attribute-cooperated convolutional neural network for remote sensing image classification. En: IEEE Transactions on Geoscience and Remote Sensing 58 (12) (2020), p. 8358–8371spa
dc.relation.referencesZhao, Z. ; Li, J. ; Luo, Z. ; Li, J. ; Chen, C.: Remote sensing image scene classification based on an enhanced attention module. En: IEEE Geoscience and Remote Sensing Letters 18 (11) (2021), p. 1926–1930spa
dc.relation.referencesZhong, Y. ; Fei, F. ; Liu, Y. ; Zhao, B. ; Jiao, H. ; Zhang, L.: SatCNN: satellite image dataset classification using agile convolutional neural networks. En: Remote Sensing Letters 8 (2) (2017), p. 136–145spa
dc.relation.referencesZhou, S. ; Zhou, Y. ; Liu, B.: Using siamese capsule networks for remote sensing scene classification. En: Remote Sensing Letters 11 (8) (2020), p. 757–766spa
dc.relation.referencesZhu, Y. ; Geis, C. ; So, E. ; Jin, Y.: Multitemporal relearning with convolutional LSTM models for land use classification. En: IEEE Journal of Selected Topics in Applied Earth Observasions and Remote Sensing 14 (2021), p. 3251–3265spa
dc.rights.accessrightsinfo:eu-repo/semantics/openAccessspa
dc.rights.licenseAtribución-NoComercial 4.0 Internacionalspa
dc.rights.urihttp://creativecommons.org/licenses/by-nc/4.0/spa
dc.subject.agrovocAprendizaje automático
dc.subject.agrovocmachine learning
dc.subject.ddc630 - Agricultura y tecnologías relacionadas::633 - Cultivos de campo y de plantaciónspa
dc.subject.ddc620 - Ingeniería y operaciones afines::629 - Otras ramas de la ingenieríaspa
dc.subject.proposalImágenes satelitalesspa
dc.subject.proposalZonas agrícolas en Colombiaspa
dc.subject.proposalAprendizaje por transferenciaspa
dc.subject.proposalRedes neuronales profundasspa
dc.subject.proposalSatellite imageseng
dc.subject.proposalAgricultural zones in Colombiaeng
dc.subject.proposalTransfer learningeng
dc.subject.proposalDeep neural networkseng
dc.subject.unescoSistema de información geográficaspa
dc.subject.unescoGeographical information systemseng
dc.subject.unescoZona ruralspa
dc.subject.unescoRural areaseng
dc.titleClasificación de zonas agrícolas en Colombia por medio de imágenes satelitales con redes neuronales profundasspa
dc.title.translatedClassification of agricultural areas in Colombia through satellite images with deep neural networkseng
dc.typeTrabajo de grado - Maestríaspa
dc.type.coarhttp://purl.org/coar/resource_type/c_bdccspa
dc.type.coarversionhttp://purl.org/coar/version/c_ab4af688f83e57aaspa
dc.type.contentTextspa
dc.type.driverinfo:eu-repo/semantics/masterThesisspa
dc.type.redcolhttp://purl.org/redcol/resource_type/TMspa
dc.type.versioninfo:eu-repo/semantics/acceptedVersionspa
dcterms.audience.professionaldevelopmentEstudiantesspa
dcterms.audience.professionaldevelopmentInvestigadoresspa
dcterms.audience.professionaldevelopmentMaestrosspa
dcterms.audience.professionaldevelopmentPúblico generalspa
dcterms.audience.professionaldevelopmentResponsables políticosspa
oaire.accessrightshttp://purl.org/coar/access_right/c_abf2spa

Archivos

Bloque original

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
1018486698.2024.pdf
Tamaño:
8.9 MB
Formato:
Adobe Portable Document Format
Descripción:
Tesis de Maestría en Ingeniería - Ingeniería de Sistemas y Computación

Bloque de licencias

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
license.txt
Tamaño:
5.74 KB
Formato:
Item-specific license agreed upon to submission
Descripción: