Terrain Super-resolution through Aerial Imagery and Fully Convolutional Networks

Simple item page
dc.contributor.authorArgudo, Oscaren_US
dc.contributor.authorChica, Antonioen_US
dc.contributor.authorAndujar, Carlosen_US
dc.contributor.editorGutierrez, Diego and Sheffer, Allaen_US
dc.date.accessioned2018-04-14T18:23:09Z
dc.date.available2018-04-14T18:23:09Z
dc.date.issued2018
dc.description.abstractDespite recent advances in surveying techniques, publicly available Digital Elevation Models (DEMs) of terrains are lowresolution except for selected places on Earth. In this paper we present a new method to turn low-resolution DEMs into plausible and faithful high-resolution terrains. Unlike other approaches for terrain synthesis/amplification (fractal noise, hydraulic and thermal erosion, multi-resolution dictionaries), we benefit from high-resolution aerial images to produce highly-detailed DEMs mimicking the features of the real terrain. We explore different architectures for Fully Convolutional Neural Networks to learn upsampling patterns for DEMs from detailed training sets (high-resolution DEMs and orthophotos), yielding up to one order of magnitude more resolution. Our comparative results show that our method outperforms competing data amplification approaches in terms of elevation accuracy and terrain plausibility.en_US
dc.description.number2
dc.description.sectionheadersGeometry Learning
dc.description.seriesinformationComputer Graphics Forum
dc.description.volume37
dc.identifier.doi10.1111/cgf.13345
dc.identifier.issn1467-8659
dc.identifier.pages101-110
dc.identifier.urihttps://doi.org/10.1111/cgf.13345
dc.identifier.urihttps://diglib.eg.org:443/handle/10.1111/cgf13345
dc.publisherThe Eurographics Association and John Wiley & Sons Ltd.en_US
dc.subjectComputing methodologies
dc.subjectImage processing
dc.subjectShape modeling
dc.titleTerrain Super-resolution through Aerial Imagery and Fully Convolutional Networksen_US
Files
Collections
37-Issue 2