{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,12]],"date-time":"2026-03-12T01:10:55Z","timestamp":1773277855979,"version":"3.50.1"},"reference-count":66,"publisher":"MDPI AG","issue":"2","license":[{"start":{"date-parts":[[2021,1,17]],"date-time":"2021-01-17T00:00:00Z","timestamp":1610841600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Remote Sensing"],"abstract":"<jats:p>Generative adversarial networks (GANs) are a type of neural network that are characterized by their unique construction and training process. Utilizing the concept of the latent space and exploiting the results of a duel between different GAN components opens up interesting opportunities for computer vision (CV) activities, such as image inpainting, style transfer, or even generative art. GANs have great potential to support aerial and satellite image interpretation activities. Carefully crafting a GAN and applying it to a high-quality dataset can result in nontrivial feature enrichment. In this study, we have designed and tested an unsupervised procedure capable of engineering new features by shifting real orthophotos into the GAN\u2019s underlying latent space. Latent vectors are a low-dimensional representation of the orthophoto patches that hold information about the strength, occurrence, and interaction between spatial features discovered during the network training. Latent vectors were combined with geographical coordinates to bind them to their original location in the orthophoto. In consequence, it was possible to describe the whole research area as a set of latent vectors and perform further spatial analysis not on RGB images but on their lower-dimensional representation. To accomplish this goal, a modified version of the big bidirectional generative adversarial network (BigBiGAN) has been trained on a fine-tailored orthophoto imagery dataset covering the area of the Pilica River region in Poland. Trained models, precisely the generator and encoder, have been utilized during the processes of model quality assurance and feature engineering, respectively. Quality assurance was performed by measuring model reconstruction capabilities and by manually verifying artificial images produced by the generator. The feature engineering use case, on the other hand, has been presented in a real research scenario that involved splitting the orthophoto into a set of patches, encoding the patch set into the GAN latent space, grouping similar patches latent codes by utilizing hierarchical clustering, and producing a segmentation map of the orthophoto.<\/jats:p>","DOI":"10.3390\/rs13020306","type":"journal-article","created":{"date-parts":[[2021,1,20]],"date-time":"2021-01-20T03:34:25Z","timestamp":1611113665000},"page":"306","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":8,"title":["Aerial Imagery Feature Engineering Using Bidirectional Generative Adversarial Networks: A Case Study of the Pilica River Region, Poland"],"prefix":"10.3390","volume":"13","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-8229-9661","authenticated-orcid":false,"given":"Maciej","family":"Adamiak","sequence":"first","affiliation":[{"name":"SoftwareMill, 02-791 Warsaw, Poland"}]},{"given":"Krzysztof","family":"B\u0119dkowski","sequence":"additional","affiliation":[{"name":"Faculty of Geographical Sciences, Institute of Urban Geography, Tourism and Geoinformation, University of Lodz, 90-139 \u0141\u00f3d\u017a, Poland"}]},{"given":"Anna","family":"Majchrowska","sequence":"additional","affiliation":[{"name":"Department of Physical Geography, Faculty of Geographical Sciences, University of Lodz, 90-139 \u0141\u00f3d\u017a, Poland"}]}],"member":"1968","published-online":{"date-parts":[[2021,1,17]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"778","DOI":"10.1109\/LGRS.2017.2681128","article-title":"Deep Learning Classification of Land Cover and Crop Types Using Remote Sensing Data","volume":"14","author":"Kussul","year":"2017","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_2","unstructured":"(2020, October 24). Landsat 8\u2014Landsat Science, Available online: https:\/\/landsat.gsfc.nasa.gov\/landsat-8\/."},{"key":"ref_3","unstructured":"(2020, October 24). Sentinel-2\u2014ESA Operational EO Missions\u2014Earth Online\u2014ESA. Available online: https:\/\/earth.esa.int\/web\/guest\/missions\/esa-operational-eo-missions\/sentinel-2."},{"key":"ref_4","unstructured":"(2020, October 24). Pleiades\u2014eoPortal Directory\u2014Satellite Missions. Available online: https:\/\/earth.esa.int\/web\/eoportal\/satellite-missions\/p\/pleiades."},{"key":"ref_5","unstructured":"(2020, October 24). Dziennik Ustaw 2020 r. poz. 1086, Available online: https:\/\/www.dziennikustaw.gov.pl\/DU\/2020\/1086."},{"key":"ref_6","unstructured":"Ol\u0119dzki, J.R. (1992). Geographical Conditions of the Diversity of the Satellite Image of Poland and Its Division into Photomorphic Units, Wydawnictwa Uniwersytetu Warszawskiego."},{"key":"ref_7","first-page":"302","article-title":"Geographical regions of Poland","volume":"38","year":"2001","journal-title":"Teledetekcja \u015arodowiska"},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Haralick, R.M., Shanmugam, K., and Dinstein, I. (1973). Textural Features for Image Classification. IEEE Trans. Syst. Man Cybern., 610\u2013621.","DOI":"10.1109\/TSMC.1973.4309314"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"100","DOI":"10.1016\/S0734-189X(85)90153-7","article-title":"Image segmentation techniques","volume":"29","author":"Haralick","year":"1985","journal-title":"Comput. Vis. Graph. Image Process."},{"key":"ref_10","unstructured":"(2020, November 12). eCognition | Trimble Geospatial. Available online: https:\/\/geospatial.trimble.com\/products-and-solutions\/ecognition."},{"key":"ref_11","first-page":"227","article-title":"Application of the object-based image analysis of VHR satellite images in land-use classification","volume":"4","year":"2006","journal-title":"Rocz. Geomatyki-Ann. Geomat."},{"key":"ref_12","unstructured":"Kondracki, J. (1977). Physical and Geographic Regions of Poland, Wydawnictwa Uniwersytetu Warszawskiego."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"143","DOI":"10.7163\/GPol.0115","article-title":"Physico-geographical mesoregions of Poland: Verification and adjustment of boundaries on the basis of contemporary spatial data","volume":"91","author":"Solon","year":"2018","journal-title":"Geogr. Pol."},{"key":"ref_14","first-page":"139","article-title":"Identification of land cover and land use forms on landsat ETM+ satellite image using the method of object-oriented classification","volume":"4","year":"2006","journal-title":"Rocz. Geomatyki-Ann. Geomat."},{"key":"ref_15","first-page":"63","article-title":"Comparison of object-oriented classification to traditional pixel-based classification with reference to automation of the process of land cover and land use data base creation","volume":"5","year":"2007","journal-title":"Rocz. Geomatyki-Ann. Geomat."},{"key":"ref_16","first-page":"69","article-title":"Application of Region Growing procedure to meadow classification based on Landsat ETM+ images","volume":"3","year":"2005","journal-title":"Rocz. Geomatyki-Ann. Geomat."},{"key":"ref_17","first-page":"385","article-title":"Application of structural features in the Landsat ETM+ image in object classification of landscape-vegetation complexes","volume":"17a","year":"2007","journal-title":"Archiwum Fotogram. Kartografii i Teledetekcji"},{"key":"ref_18","first-page":"148","article-title":"Multiscale object-based image analysis\u2014A key to the hierarchical organisation of landscapes","volume":"23","author":"Lang","year":"2004","journal-title":"Ekol\u00f3gia"},{"key":"ref_19","first-page":"37","article-title":"Object-based analysis as a method of improving classification quality","volume":"4","author":"Adamczyk","year":"2006","journal-title":"Rocz. Geomatyki-Ann. Geomat."},{"key":"ref_20","first-page":"9","article-title":"The effect of scaling methods on the calculation of environmental indices","volume":"17","author":"Adamczyk","year":"2013","journal-title":"Ecol. Quest."},{"key":"ref_21","first-page":"21","article-title":"Application of satellite images and GIS technology for updating of soil-landscape units\u2014Examples for small scales mapping","volume":"21","author":"Chmiel","year":"2010","journal-title":"Archiwum Fotogram. Kartografii i Teledetekcji"},{"key":"ref_22","first-page":"87","article-title":"Application of the geodiversity index for defining the relief\u2019s diversity based on the example of the Struga Toru\u0144ska representative basin, Che\u0142mno Lakeland","volume":"33","author":"Kot","year":"2012","journal-title":"Probl. Ekol. Kraj."},{"key":"ref_23","first-page":"193","article-title":"Assessment of diversity of landscape on the basis of analysis of spatial structure of vegetation","volume":"185","author":"Solon","year":"2002","journal-title":"Pr. Geogr."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"699","DOI":"10.1080\/014311697218700","article-title":"Introduction Neural networks in remote sensing","volume":"18","author":"Atkinson","year":"1997","journal-title":"Int. J. Remote Sens."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"2094","DOI":"10.1109\/JSTARS.2014.2329330","article-title":"Deep Learning-Based Classification of Hyperspectral Data","volume":"7","author":"Chen","year":"2014","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"2448","DOI":"10.1109\/LGRS.2015.2483680","article-title":"Multiview Deep Learning for Land-Use Classification","volume":"12","author":"Luus","year":"2015","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Adamiak, M., Biczkowski, M., Le\u015bniewska-Napiera\u0142a, K., Nalej, M., and Napiera\u0142a, T. (2020). Impairing Land Registry: Social, Demographic, and Economic Determinants of Forest Classification Errors. Remote Sens., 12.","DOI":"10.3390\/rs12162628"},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Cabezas, M., Kentsch, S., Tomhave, L., Gross, J., Caceres, M.L.L., and Diez, Y. (2020). Detection of Invasive Species in Wetlands: Practical DL with Heavily Imbalanced Data. Remote Sens., 12.","DOI":"10.3390\/rs12203431"},{"key":"ref_29","unstructured":"Sobczak, M., Folbrier, A., Koz\u0142owska, A., Pabjanek, P., Wrzesien, M., and Zagajewski, B. (2005). Assessment of the potential of hyperspectral data and techniques for mountain vegetation analysis. Imaging Spectroscopy: New Quality in Environmental Studies, EARSeL & Warsaw University."},{"key":"ref_30","first-page":"1","article-title":"Assessment of neural networks and Imaging Spectroscopy for vegetation classification of the High Tatras","volume":"43","author":"Zagajewski","year":"2010","journal-title":"Teledetekcja \u015arodowiska"},{"key":"ref_31","first-page":"5","article-title":"Applying neural networks to urban area classification in satellite images","volume":"1","author":"Iwaniak","year":"2002","journal-title":"Acta Scientiarum Polonorum. Geodesia et Descriptio Terrarum"},{"key":"ref_32","first-page":"361","article-title":"Supervised road detection using machine learning methodology","volume":"16","author":"Krawiec","year":"2006","journal-title":"Archiwum Fotogram. Kartografii Teledetekcji"},{"key":"ref_33","unstructured":"Goodfellow, I.J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., and Bengio, Y. (2014). Generative Adversarial Networks. arXiv."},{"key":"ref_34","unstructured":"Salimans, T., Goodfellow, I., Zaremba, W., Cheung, V., Radford, A., and Chen, X. (2016). Improved Techniques for Training GANs. arXiv."},{"key":"ref_35","doi-asserted-by":"crossref","unstructured":"Dong, R., Li, C., Fu, H., Wang, J., Li, W., Yao, Y., Gan, L., Yu, L., and Gong, P. (2020). Improving 3-m Resolution Land Cover Mapping through Efficient Learning from an Imperfect 10-m Resolution Map. Remote Sens., 12.","DOI":"10.3390\/rs12091418"},{"key":"ref_36","doi-asserted-by":"crossref","unstructured":"Burdziakowski, P. (2020). A Novel Method for the Deblurring of Photogrammetric Images Using Conditional Generative Adversarial Networks. Remote Sens., 12.","DOI":"10.3390\/rs12162586"},{"key":"ref_37","doi-asserted-by":"crossref","unstructured":"Zhao, W., Chen, X., Chen, J., and Qu, Y. (2020). Sample Generation with Self-Attention Generative Adversarial Adaptation Network (SaGAAN) for Hyperspectral Image Classification. Remote Sens., 12.","DOI":"10.3390\/rs12050843"},{"key":"ref_38","unstructured":"Donahue, J., Kr\u00e4henb\u00fchl, P., and Darrell, T. (2017). Adversarial Feature Learning. arXiv."},{"key":"ref_39","unstructured":"\u017bela\u017aniewicz, A., Aleksandrowski, P., Bu\u0142a, Z., Karnkowski, P.H., Konon, A., \u015al\u0105czka, A., \u017baba, J., and \u017bytko, K. (2011). Tectonic Regionalization of Poland, Komitet Nauk Geologicznych PAN."},{"key":"ref_40","unstructured":"Head Office of Geodesy and Cartography (2021, January 09). Data for the Numerical Terrain Model with a Grid Interval of at Least 100 m, Available online: http:\/\/www.gugik.gov.pl\/pzgik\/dane-bez-oplat\/dane-dotyczace-numerycznego-modelu-terenu-o-interwale-siatki-co-najmniej-100-m-nmt_100."},{"key":"ref_41","first-page":"299","article-title":"Ecological aspects of land use changes in chosen types of natural landscapes in Central Poland","volume":"21","author":"Krysiak","year":"2008","journal-title":"Probl. Ekol. Kraj."},{"key":"ref_42","unstructured":"(2020, November 11). Geoportal.gov.pl, Available online: http:\/\/geoportal.gov.pl."},{"key":"ref_43","doi-asserted-by":"crossref","first-page":"78","DOI":"10.1145\/2347736.2347755","article-title":"A few useful things to know about machine learning","volume":"55","author":"Domingos","year":"2012","journal-title":"Commun. ACM"},{"key":"ref_44","unstructured":"Donahue, J., and Simonyan, K. (2019). Large Scale Adversarial Representation Learning. arXiv."},{"key":"ref_45","doi-asserted-by":"crossref","unstructured":"Mukherjee, S., Asnani, H., Lin, E., and Kannan, S. (2019). ClusterGAN: Latent Space Clustering in Generative Adversarial Networks. arXiv.","DOI":"10.1609\/aaai.v33i01.33014610"},{"key":"ref_46","unstructured":"(2020, October 24). 2.3. Clustering\u2014Scikit-Learn 0.23.2 Documentation. Available online: https:\/\/scikit-learn.org\/stable\/modules\/clustering.html#hierarchical-clustering."},{"key":"ref_47","doi-asserted-by":"crossref","first-page":"236","DOI":"10.1080\/01621459.1963.10500845","article-title":"Hierarchical Grouping to Optimize an Objective Function","volume":"58","author":"Ward","year":"1963","journal-title":"J. Am. Stat. Assoc."},{"key":"ref_48","unstructured":"Radford, A., Metz, L., and Chintala, S. (2016). Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks. arXiv."},{"key":"ref_49","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S., and Sun, J. (2015). Deep Residual Learning for Image Recognition. arXiv.","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref_50","doi-asserted-by":"crossref","unstructured":"Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., and Rabinovich, A. (2014). Going Deeper with Convolutions. arXiv.","DOI":"10.1109\/CVPR.2015.7298594"},{"key":"ref_51","unstructured":"Brock, A., Donahue, J., and Simonyan, K. (2019). Large Scale GAN Training for High Fidelity Natural Image Synthesis. arXiv."},{"key":"ref_52","doi-asserted-by":"crossref","first-page":"450","DOI":"10.1016\/0047-259X(82)90077-X","article-title":"The Fr\u00e9chet distance between multivariate normal distributions","volume":"12","author":"Landau","year":"1982","journal-title":"J. Multivar. Anal."},{"key":"ref_53","unstructured":"Zhou, S., Gordon, M.L., Krishna, R., Narcomey, A., Fei-Fei, L., and Bernstein, M.S. (2019). HYPE: A Benchmark for Human eYe Perceptual Evaluation of Generative Models. arXiv."},{"key":"ref_54","doi-asserted-by":"crossref","first-page":"600","DOI":"10.1109\/TIP.2003.819861","article-title":"Image quality assessment: From error visibility to structural similarity","volume":"13","author":"Wang","year":"2004","journal-title":"IEEE Trans. Image Process."},{"key":"ref_55","doi-asserted-by":"crossref","unstructured":"Korhonen, J., and You, J. (2012, January 5\u20137). Peak signal-to-noise ratio revisited: Is simple beautiful?. Proceedings of the 2012 Fourth International Workshop on Quality of Multimedia Experience, Yarra Valley, Australia.","DOI":"10.1109\/QoMEX.2012.6263880"},{"key":"ref_56","unstructured":"(2020, November 11). Head Office of Geodesy and Cartography Integrated Copies of Databases of Topographic Objects, Available online: http:\/\/www.gugik.gov.pl\/pzgik\/zamow-dane\/baza-danych-obiektow-topograficznych-bdot-10k."},{"key":"ref_57","unstructured":"(2020, October 24). GeoPandas 0.8.0\u2014GeoPandas 0.8.0 Documentation. Available online: https:\/\/geopandas.org\/."},{"key":"ref_58","unstructured":"(2020, October 24). sklearn.neighbors.KNeighborsClassifier\u2014Scikit-Learn 0.23.2 Documentation. Available online: https:\/\/scikit-learn.org\/stable\/modules\/generated\/sklearn.neighbors.KNeighborsClassifier.html."},{"key":"ref_59","doi-asserted-by":"crossref","first-page":"307","DOI":"10.1561\/2200000056","article-title":"An Introduction to Variational Autoencoders","volume":"12","author":"Kingma","year":"2019","journal-title":"FNT Mach. Learn."},{"key":"ref_60","unstructured":"Frogner, C., Zhang, C., Mobahi, H., Araya-Polo, M., and Poggio, T. (2015). Learning with a Wasserstein Loss. arXiv."},{"key":"ref_61","unstructured":"Gulrajani, I., Ahmed, F., Arjovsky, M., Dumoulin, V., and Courville, A. (2017). Improved Training of Wasserstein GANs. arXiv."},{"key":"ref_62","unstructured":"Miyato, T., Kataoka, T., Koyama, M., and Yoshida, Y. (2018). Spectral Normalization for Generative Adversarial Networks. arXiv."},{"key":"ref_63","doi-asserted-by":"crossref","unstructured":"Deora, P., Vasudeva, B., Bhattacharya, S., and Pradhan, P.M. (2020). Structure Preserving Compressive Sensing MRI Reconstruction using Generative Adversarial Networks. arXiv.","DOI":"10.1109\/CVPRW50498.2020.00269"},{"key":"ref_64","doi-asserted-by":"crossref","unstructured":"Thanh-Tung, H., and Tran, T. (2020). On Catastrophic Forgetting and Mode Collapse in Generative Adversarial Networks. arXiv.","DOI":"10.1109\/IJCNN48605.2020.9207181"},{"key":"ref_65","unstructured":"(2020, October 26). ImageNet: A Large-Scale Hierarchical Image Database\u2014IEEE Conference Publication. Available online: https:\/\/ieeexplore.ieee.org\/document\/5206848."},{"key":"ref_66","unstructured":"Forslund, L. (2020, November 11). CLC 2018\u2014Copernicus Land Monitoring Service. Available online: https:\/\/land.copernicus.eu\/pan-european\/corine-land-cover\/clc2018."}],"container-title":["Remote Sensing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2072-4292\/13\/2\/306\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T05:12:05Z","timestamp":1760159525000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2072-4292\/13\/2\/306"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,1,17]]},"references-count":66,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2021,1]]}},"alternative-id":["rs13020306"],"URL":"https:\/\/doi.org\/10.3390\/rs13020306","relation":{},"ISSN":["2072-4292"],"issn-type":[{"value":"2072-4292","type":"electronic"}],"subject":[],"published":{"date-parts":[[2021,1,17]]}}}