{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,22]],"date-time":"2026-03-22T00:56:02Z","timestamp":1774140962180,"version":"3.50.1"},"reference-count":39,"publisher":"MDPI AG","issue":"11","license":[{"start":{"date-parts":[[2023,5,26]],"date-time":"2023-05-26T00:00:00Z","timestamp":1685059200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Ministry of Finance in Finland (Valtiovarainministeri\u00f6)"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Remote Sensing"],"abstract":"<jats:p>Vector datasets of small watercourses, such as rivulets, streams, and ditches, are important for many visualization and analysis use cases. Mapping small watercourses with traditional methods is laborious and costly. Convolutional neural networks (CNNs) are state-of-the-art computer vision methods that have been shown to be effective for extracting geospatial features, including small watercourses, from LiDAR point clouds, digital elevation models (DEMs), and aerial images. However, the cause of the false predictions by machine-learning models is often not thoroughly explored, and thus the impact of the results on the process of producing accurate datasets is not well understood. We digitized a highly accurate and complete dataset of small watercourses from a study area in Finland. We then developed a process based on a CNN that can be used to extract small watercourses from DEMs. We tested and validated the performance of the network with different input data layers, and their combinations to determine the best-performing layer. We analyzed the false predictions to gain an understanding of their nature. We also trained models where watercourses with high levels of uncertainty were removed from the training sets and compared the results to training models with all watercourses in the training set. The results show that the DEM was the best-performing layer and that combinations of layers provided worse results. Major causes of false predictions were shown to be boundary errors with an offset between the prediction and labeled data, as well as errors of omission by watercourses with high levels of uncertainty. Removing features with the highest level of uncertainty from the labeled dataset increased the overall f1-score but reduced the recall of the remaining features. Additional research is required to determine if the results remain similar to other CNN methods.<\/jats:p>","DOI":"10.3390\/rs15112776","type":"journal-article","created":{"date-parts":[[2023,5,27]],"date-time":"2023-05-27T16:17:33Z","timestamp":1685204253000},"page":"2776","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":8,"title":["Mapping Small Watercourses from DEMs with Deep Learning\u2014Exploring the Causes of False Predictions"],"prefix":"10.3390","volume":"15","author":[{"given":"Christian","family":"Koski","sequence":"first","affiliation":[{"name":"Finnish Geospatial Research Institute (FGI) in the National Land Survey of Finland (NLS), Vuorimiehentie, 02150 Espoo, Finland"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7025-4570","authenticated-orcid":false,"given":"Pyry","family":"Kettunen","sequence":"additional","affiliation":[{"name":"Finnish Geospatial Research Institute (FGI) in the National Land Survey of Finland (NLS), Vuorimiehentie, 02150 Espoo, Finland"}]},{"given":"Justus","family":"Poutanen","sequence":"additional","affiliation":[{"name":"Finnish Geospatial Research Institute (FGI) in the National Land Survey of Finland (NLS), Vuorimiehentie, 02150 Espoo, Finland"}]},{"given":"Lingli","family":"Zhu","sequence":"additional","affiliation":[{"name":"Finnish Geospatial Research Institute (FGI) in the National Land Survey of Finland (NLS), Vuorimiehentie, 02150 Espoo, Finland"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3381-9763","authenticated-orcid":false,"given":"Juha","family":"Oksanen","sequence":"additional","affiliation":[{"name":"Finnish Geospatial Research Institute (FGI) in the National Land Survey of Finland (NLS), Vuorimiehentie, 02150 Espoo, Finland"}]}],"member":"1968","published-online":{"date-parts":[[2023,5,26]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"1026","DOI":"10.1002\/hyp.11472","article-title":"Extracting drainage networks and their connectivity using LiDAR data","volume":"32","author":"Roelens","year":"2018","journal-title":"Hydrol. Process."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"171","DOI":"10.1016\/j.cageo.2017.06.003","article-title":"Automatic identification of watercourses in flat and engineered landscapes by computing the skeleton of a LiDAR point cloud","volume":"106","author":"Broersen","year":"2017","journal-title":"Comput. Geosci."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"18","DOI":"10.3390\/s18010018","article-title":"Comparison of random forest, k-nearest neighbor, and support vector machine classifiers for land cover classification using Sentinel-2 imagery","volume":"18","author":"Noi","year":"2017","journal-title":"Sensors"},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"La Salandra, M., Colacicco, R., Dellino, P., and Capolongo, D. (2023). An Effective Approach for Automatic River Features Extraction Using High-Resolution UAV Imagery. Drones, 7.","DOI":"10.3390\/drones7020070"},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"Stanislawski, L.V., Shavers, E.J., Wang, S., Jiang, Z., Usery, E.L., Moak, E., Duffy, A., and Schott, J. (2021). Extensibility of U-Net neural network model for hydrographic feature extraction and implications for hydrologic modeling. Remote Sens., 13.","DOI":"10.3390\/rs13122368"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"354","DOI":"10.1016\/j.patcog.2017.10.013","article-title":"Recent advances in convolutional neural networks","volume":"77","author":"Gu","year":"2018","journal-title":"Pattern Recognit."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"6999","DOI":"10.1109\/TNNLS.2021.3084827","article-title":"A survey of convolutional neural networks: Analysis, applications, and prospects","volume":"33","author":"Li","year":"2022","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"155","DOI":"10.1109\/LGRS.2018.2867736","article-title":"Building extraction from LiDAR data applying deep convolutional neural networks","volume":"16","author":"Maltezos","year":"2018","journal-title":"IEEE Geosci. Remote Sens. Lett."},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Shao, Z., Tang, P., Wang, Z., Saleem, N., Yam, S., and Sommai, C. (2020). BRRNet: A fully convolutional neural network for automatic building extraction from high-resolution remote sensing images. Remote Sens., 12.","DOI":"10.3390\/rs12061050"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"2611","DOI":"10.1109\/JSTARS.2021.3058097","article-title":"Attention-gate-based encoder\u2013decoder network for automatical building extraction","volume":"14","author":"Deng","year":"2021","journal-title":"IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"2043","DOI":"10.1109\/TGRS.2018.2870871","article-title":"RoadNet: Learning to comprehensively analyze road networks in complex urban scenes from high-resolution remotely sensed images","volume":"57","author":"Liu","year":"2018","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Xu, Z., Shen, Z., Li, Y., Xia, L., Wang, H., Li, S., Jiao, S., and Lei, Y. (2020). Road extraction in mountainous regions from high-resolution images based on DSDNet and terrain optimization. Remote Sens., 13.","DOI":"10.3390\/rs13010090"},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"1151","DOI":"10.1080\/15481603.2021.1972713","article-title":"RoadVecNet: A new approach for simultaneous road network segmentation and vectorization from aerial and google earth imagery in a complex urban set-up","volume":"58","author":"Abdollahi","year":"2021","journal-title":"GIScience Remote Sens."},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Ghorbanzadeh, O., Blaschke, T., Gholamnia, K., Meena, S.R., Tiede, D., and Aryal, J. (2019). Evaluation of different machine learning methods and deep-learning convolutional neural networks for landslide detection. Remote Sens., 11.","DOI":"10.3390\/rs11020196"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"65347","DOI":"10.1109\/ACCESS.2019.2917952","article-title":"DenseU-net-based semantic segmentation of small objects in urban remote sensing images","volume":"7","author":"Dong","year":"2019","journal-title":"IEEE Access"},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"104992","DOI":"10.1016\/j.envsoft.2021.104992","article-title":"An attention U-Net model for detection of fine-scale hydrologic streamlines","volume":"140","author":"Xu","year":"2021","journal-title":"Environ. Model. Softw."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"323","DOI":"10.1016\/S0734-189X(84)80011-0","article-title":"The extraction of drainage networks from digital elevation data","volume":"28","author":"Mark","year":"1984","journal-title":"Comput. Vis. Graph. Image Process."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"105135","DOI":"10.1016\/j.envsoft.2021.105135","article-title":"Deep learning-enhanced extraction of drainage networks from digital elevation models","volume":"144","author":"Mao","year":"2021","journal-title":"Environ. Model. Softw."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"843","DOI":"10.1002\/(SICI)1099-1085(199805)12:6<843::AID-HYP658>3.0.CO;2-R","article-title":"The treatment of flat areas and depressions in automated drainage analysis of raster digital elevation models","volume":"12","author":"Martz","year":"1998","journal-title":"Hydrol. Process."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"3489","DOI":"10.1080\/01431160701469057","article-title":"Agrarian landscapes linear features detection from LiDAR: Application to artificial drainage networks","volume":"29","author":"Bailly","year":"2008","journal-title":"Int. J. Remote Sens."},{"key":"ref_21","first-page":"853","article-title":"A spatial stochastic algorithm to reconstruct artificial drainage networks from incomplete network delineations","volume":"13","author":"Bailly","year":"2011","journal-title":"Int. J. Appl. Earth Obs. Geoinf."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"541","DOI":"10.1002\/hyp.9224","article-title":"Drainage network detection and assessment of network storage capacity in agrarian landscape","volume":"27","author":"Cazorzi","year":"2013","journal-title":"Hydrol. Process."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"1254","DOI":"10.1029\/2009JF001254","article-title":"A geometric framework for channel network extraction from lidar: Nonlinear diffusion and geodesic paths","volume":"115","author":"Passalacqua","year":"2010","journal-title":"J. Geophys. Res. Earth Surf."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"W03528","DOI":"10.1029\/2011WR010958","article-title":"Automatic geomorphic feature extraction from lidar in flat and engineered landscapes","volume":"48","author":"Passalacqua","year":"2012","journal-title":"Water Resour. Res."},{"key":"ref_25","first-page":"38","article-title":"Undrained peatland areas disturbed by surrounding drainage: A large scale GIS analysis in Finland with a special focus on aapa mires","volume":"24","author":"Sallinen","year":"2019","journal-title":"Mires Peat"},{"key":"ref_26","unstructured":"(2022, November 08). Utilised Agricultural Area 2022 (Provisional). Available online: https:\/\/www.luke.fi\/en\/statistics\/utilised-agricultural-area\/utilised-agricultural-area-2022-provisional."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"19","DOI":"10.1080\/23729333.2017.1300998","article-title":"A design of contour generation for topographic maps with adaptive DEM smoothing","volume":"3","author":"Kettunen","year":"2017","journal-title":"Int. J. Cartogr."},{"key":"ref_28","unstructured":"(2022, November 23). Laser Scanning Data. 5p. Available online: https:\/\/www.maanmittauslaitos.fi\/en\/maps-and-spatial-data\/expert-users\/product-descriptions\/laser-scanning-data-5-p."},{"key":"ref_29","first-page":"49","article-title":"Lake and river systems in Finland","volume":"23","author":"Kuusisto","year":"2006","journal-title":"Finn Environ."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"398","DOI":"10.3390\/rs3020398","article-title":"Sky-view factor as a relief visualization technique","volume":"3","author":"Kokalj","year":"2011","journal-title":"Remote Sens."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"6427","DOI":"10.3390\/rs5126427","article-title":"Openness as visualization technique for interpretative mapping of airborne lidar derived digital terrain models","volume":"5","author":"Doneus","year":"2013","journal-title":"Remote Sens."},{"key":"ref_32","unstructured":"(2022, November 23). Pytorch. Available online: https:\/\/pytorch.org."},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Ronneberger, O., Fischer, P., and Brox, T. (2015, January 5\u20139). U-net: Convolutional networks for biomedical image segmentation. Proceedings of the 18th International Conference on Medical Image Computing and Computer-Assisted Intervention, Munich, Germany.","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"ref_34","unstructured":"Kingma, D.P., and Ba, J. (2015, January 7\u20139). Adam: A method for stochastic optimization. Proceedings of the 3rd International Conference on Learning Representations, San Diego, CA, USA."},{"key":"ref_35","doi-asserted-by":"crossref","unstructured":"Mnih, V., and Hinton, G.E. (2010, January 5\u201311). Learning to detect roads in high-resolution aerial images. Proceedings of the 11th European Conference on Computer Vision, Heraklion, Greece.","DOI":"10.1007\/978-3-642-15567-3_16"},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"291","DOI":"10.1080\/13658816.2018.1530353","article-title":"Automatic determination of stream networks from DEMs by using road network data to locate culverts","volume":"33","author":"Oksanen","year":"2019","journal-title":"Int. J. Geogr. Inf. Sci."},{"key":"ref_37","first-page":"102777","article-title":"An effective superpixel-based graph convolutional network for small waterbody extraction from remotely sensed imagery","volume":"109","author":"Shi","year":"2022","journal-title":"Int. J. Appl. Earth Obs. Geoinf."},{"key":"ref_38","first-page":"103159","article-title":"A comparative study of loss functions for road segmentation in remotely sensed road datasets","volume":"116","author":"Xu","year":"2023","journal-title":"Int. J. Appl. Earth Obs. Geoinf."},{"key":"ref_39","doi-asserted-by":"crossref","unstructured":"\u015aliwi\u0144ski, D., Konieczna, A., and Roman, K. (2022). Geostatistical Resampling of LiDAR-Derived DEM in Wide Resolution Range for Modelling in SWAT: A Case Study of Zg\u0142owi\u0105czka River (Poland). Remote Sens., 14.","DOI":"10.3390\/rs14051281"}],"container-title":["Remote Sensing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2072-4292\/15\/11\/2776\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T19:42:52Z","timestamp":1760125372000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2072-4292\/15\/11\/2776"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,5,26]]},"references-count":39,"journal-issue":{"issue":"11","published-online":{"date-parts":[[2023,6]]}},"alternative-id":["rs15112776"],"URL":"https:\/\/doi.org\/10.3390\/rs15112776","relation":{},"ISSN":["2072-4292"],"issn-type":[{"value":"2072-4292","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,5,26]]}}}