KIT | KIT-Bibliothek | Impressum | Datenschutz
Suche im KITopen-Katalog

Repository KITopen

MobimaSkidRoads Dataset

Michiels, Lukas ORCID iD icon 1; Becker, Simon 1; Beiser, Sebastian ORCID iD icon 1; Geiger, Chris; Graf, Marina ORCID iD icon 1; Kazenwadel, Benjamin 1; Rivera Cedeño, Emilio 1; Waldschmidt, Martin 1; Westermann, Manuel 1; Geimer, Marcus 1
1 Institut für Fahrzeugsystemtechnik (FAST), Karlsruher Institut für Technologie (KIT)


radar4kit-icon
Download (RADAR4KIT)
BibTeX (UTF-8) BibTeX (ASCII) EndNote/Refer (.enw) RIS CSL-JSON ISI
Zugehörige Institution(en) am KIT Institut für Fahrzeugsystemtechnik (FAST)
Publikationstyp Forschungsdaten
Publikationsdatum 27.09.2023
Erstellungsdatum 31.08.2023
Identifikator DOI: 10.35097/1750
KITopen-ID: 1000162515
Lizenz Creative Commons Namensnennung – Weitergabe unter gleichen Bedingungen 4.0 International
Schlagwörter Instance Segmentation, Forest Dataset, Tree Objects
Liesmich

The MobimaSkidroads, consists of 293 images captured while driving on a skid or forest road with an industrial camera mounted on a vehicle. The dataset is completely in 16:9 landscape format and includes mixed and coniferous forests in summer and winter. The dataset includes mask and bounding box annotations for the five object classes:

  • Stems (0): Trunks and cutted trees with a diameter of approx. ≥10 cm
  • Trees (1): Trees with a diameter of approx. ≥10 cm
  • Stumps (2): Tree stumps
  • Obstacles (3): Non-passable objects not being subject to any of the other groups (e.g. stones, raised hides, way signs)
  • Persons (4): Complete or partially visible humans
Art der Forschungsdaten Dataset
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page