Skip to main content
zenodoopen

Orthophotos and DSMs derived from RPAS flights over the nature reserve Zwin in Flanders, Belgium

<p><strong>Study area</strong></p> <p>The Zwin is a nature reserve situated along the Belgian North Sea coast, northeast of Knokke, in the province of West-Flanders, Flanders, Belgium. The area is managed by the Flemish Agency for Nature and Forest and consists of a tidal marsh, coastal dunes with <em>Ammophila arenaria</em>, dune grasslands and/or shrub (<em>Hippophae rhamnoides</em>, <em>Salix repens</em>), and a transitional grassland zone that stretches from the inner edge of the coastal dunes into the polders.</p> <p><strong>Data collection</strong></p> <p>Data were collected by the <a href="http://www.inbo.be/en">Research Institute for Nature and Forest (INBO)</a> with a fixed wing drone Gatewing X100 in 2014 and 2015 (15 flights). RGB data were acquired using an off-the-shelf Ricoh GR Digital IV camera, with the following image bands: 1: red, 2: green, 3: blue, 4: alpha channel. CIR (color-infrared) data were acquired using a NIR-enabled Ricoh GR Digital IV camera, with the following info bands: 1: NIR, 2: red, 3: green, 4: alpha channel.</p> <p><strong>Data processing</strong></p> <p>The raw data were processed to Digital Surface Models and orthophotos by the <a href="https://vito.be/en">Flemish Institute for Technological Research (VITO)</a> in 2017. Images with coarse GPS coordinates were imported and processed in Agisoft PhotoScan Pro 1.4.x, a structure-from-motion (SfM) based photogrammetry software program. After extraction and matching of tie points, a bundle adjustment leads to a sparse point cloud and a refined set of camera position and orientation values. Ground control points (either artificially installed markers on the terrain, or other photo-identifiable points, measured on the ground with RTK GNSS) were used to further refine the camera calibration and obtain a pixel-level georeferencing accuracy. From there, a point cloud densification and classification into ground and non-ground points was performed, leading to a rasterized digital surface model (DSM) and digital terrain model (DTM). Finally, a true orthomosaic was projected onto the DTM.</p> <p><strong>Coordinate reference system</strong></p> <p>All geospatial data have the coordinate reference system <code>EPSG:31370 - Belgian Lambert 72</code>.</p> <p><strong>Files</strong></p> <ul> <li><strong>Raw flight data</strong>: images and logs collected by the drone during flight. These files are zipped per flight, with the date (<code>yyyymmdd</code>) and flight number (<code>x</code>) indicated in the file name (<code>flight_yyyymmdd_Zwin_x.zip</code>).</li> <li><strong>Processed data</strong>: Digital Surface Models (<code>filename_DSM.tif</code>) and orthophotos (<code>filename_Ortho.tif</code>) stitched together from the raw data. The included flights are indicated in the file name (e.g. 6 flights for <code>20150709_Zwin_1-3_20150710_Zwin_1-3_DSM.tif</code>).</li> <li><strong>Ground control points</strong>: fixed ground control points (GCP) were placed on 2014-04-07, coordinates of which are available in <code>GCP_20140407_Zwin_fixed.tsv</code>. These GCPs are visible (but fading over time) in all orthophotos except <code>20151012_Zwin_1-4_Ortho.tif</code> which covers a different area. Additional temporary GCPs were placed on 2014-04-07, 2014-04-10 and 2015-07-09 (visible in orthophotos of those dates), coordinates of which are available in the respective <code>GCP_yyyymmdd_Zwin.tsv</code> file.</li> </ul> <p><strong>Cloud Optimized GeoTIFF</strong></p> <p>The most efficient way to explore the processed data is by loading the <a href="https://www.cogeo.org/">Cloud Optimized GeoTIFFs</a> we created for each processed file. Copy one of the file URLs below and follow e.g. the <a href="https://www.cogeo.org/qgis-tutorial.html">QGIS tutorial</a> to load this type of file.</p> <ul> <li><code>http://s3-eu-west-1.amazonaws.com/lw-remote-sensing/cogeo/20140407_Zwin_1-2_DSM.tif</code></li> <li><code>http://s3-eu-west-1.amazonaws.com/lw-remote-sensing/cogeo/20140407_Zwin_1-2_Ortho.tif</code>&nbsp;CIR</li> <li><code>http://s3-eu-west-1.amazonaws.com/lw-remote-sensing/cogeo/20140410_Zwin_1-3_DSM.tif</code></li> <li><code>http://s3-eu-west-1.amazonaws.com/lw-remote-sensing/cogeo/20140410_Zwin_1-3_Ortho.tif</code>&nbsp;CIR</li> <li><code>http://s3-eu-west-1.amazonaws.com/lw-remote-sensing/cogeo/20150709_Zwin_1-3_20150710_Zwin_1-3_DSM.tif</code></li> <li><code>http://s3-eu-west-1.amazonaws.com/lw-remote-sensing/cogeo/20150709_Zwin_1-3_20150710_Zwin_1-3_Ortho.tif</code> RGB</li> <li><code>http://s3-eu-west-1.amazonaws.com/lw-remote-sensing/cogeo/20151012_Zwin_1-4_DSM.tif</code></li> <li><code>http://s3-eu-west-1.amazonaws.com/lw-remote-sensing/cogeo/20151012_Zwin_1-4_Ortho.tif</code> RGB</li> </ul> <p>See <a href="https://s3-eu-west-1.amazonaws.com/lw-remote-sensing/index.html">this page</a> for an overview of public INBO RPAS data.</p>

ShareScore

40/100

Overall dataset sharing score

Score breakdown

These five areas show where the dataset supports — or may limit — practical reuse.

Stewardship
8
Harmonization
4
Access
20
Reuse readiness
8
Engagement
0

Topics