Poster Open Access

Semi-supervised Sentinel-2 tree species detection

Vermeer, Martijn; Sørensen, Tord K.; Völgyes, David; Fantin, Daniele; Miller, Heidrun


Citation Style Language JSON Export

{
  "publisher": "Zenodo", 
  "DOI": "10.5281/zenodo.6563721", 
  "language": "eng", 
  "title": "Semi-supervised Sentinel-2 tree species detection", 
  "issued": {
    "date-parts": [
      [
        2022, 
        5, 
        19
      ]
    ]
  }, 
  "abstract": "<p>For forest management the availability of complete and up-to-date forest inventories is essential, with one of the most important parameters&nbsp;being the volumetric tree species distribution. Unfortunately, tree species mapping in Norwegian&nbsp;production forests is a time-consuming and largely&nbsp;manual process, leading to forest inventories that&nbsp;are often incomplete and/or outdated. Indeed, the&nbsp;determination of the tree species distribution is&nbsp;currently performed by a forestry expert, mainly<br>\nby visual interpretation of aerial imagery and in&nbsp;some cases lidar data. High resolution aerial imagery is available, however&nbsp; campaigns are expensive and therefore infrequent. Satellite imagery, on&nbsp;the other hand, provides dense time series, but has&nbsp;a much lower resolution. The primary goal of the&nbsp;SENTREE project is to automate the classification&nbsp;of Norwegian main production tree species (Norway spruce, Scots pine and Birch) using semantic&nbsp;segmentation networks on a fusion of aerial and&nbsp;satellite data sources.</p>", 
  "author": [
    {
      "family": "Vermeer, Martijn"
    }, 
    {
      "family": "S\u00f8rensen, Tord K."
    }, 
    {
      "family": "V\u00f6lgyes, David"
    }, 
    {
      "family": "Fantin, Daniele"
    }, 
    {
      "family": "Miller, Heidrun"
    }
  ], 
  "type": "graphic", 
  "id": "6563721"
}
52
51
views
downloads
All versions This version
Views 5252
Downloads 5151
Data volume 681.5 MB681.5 MB
Unique views 4848
Unique downloads 4141

Share

Cite as