Loading...
hidden

Mobile-Version anzeigen

Meta-Navigation

Startseite – Hochschule Luzern

Sprachwahl und wichtige Links

  • Zum Inhalt springen
  • Kontakt
  • Login
  • De
Suche starten

Hauptnavigation

Departementsnavigation

  • Technik & Architektur
  • Wirtschaft
  • Informatik
  • Soziale Arbeit
  • Design & Kunst
  • Musik

Unternavigation

  • Studium
  • Weiterbildung
  • Forschung
  • Institute
  • Über uns

Unternavigation

Breadcrumbs-Navigation

  1. Technik & Architektur Technik & Architektur
  2. Über uns Über uns
  3. Institute im Bereich Technik Institute im Bereich Technik
  4. Elektrotechnik Elektrotechnik
  5. edgeAI@T&A edgeAI@T&A
  6. Image-Dataset Generation for Supervised Learning of Space Debris Dtectors Image-Dataset Generation for Supervised Learning of Space Debris Dtectors

Image-Dataset Generation for Supervised Learning of Space Debris Dtectors Rendering to the rescue 

Insufficient data for training of machine learning models is often a challenge, and is particularly critical for space applications. Fully parametrizable and automatically labelled datasets can be generated using a rendering environment. 

Space debris is an ever growing concern for humans ambitions in space. Thousands of objects are currentlytracked to make sure no collisions with expensive hardware or even crewed spacecraft happen. Stricterlaws regarding the end-of-life of space hardware are expected to be introduced in the coming years. However, even with strictest laws, there is already a large amount of objects around earth, that will take thousands of years to reenter the earth atmosphere.This work is part of a project on active space debris removal from orbit using a dedicated space probe. This space probe is doing the final approach to a target object using a camera based navigation system. For training and evaluation of such a system, a labeled data set is needed. This work introduces a Blender based rendering pipeline for generating large training datasets for the application of space debris detection using camera data. The approach described in this work allows parameterized data set generation for a multitude of possible environmental conditions. The flexible nature of this approach allows the tool to be used also for other machine learning applications where object detection, localization and image segmentation might be of interest.Preliminary tests were made to evaluate the possibilities of such data sets. Different approaches were compared and evaluated on the basis of the constraint environment of a space probe, where size and power constraints as well as the radiation environment represent major challenges in regards to computing resources available to the localization algorithms. The evaluations suggests, that with the help of this rendering pipeline a ML approach is indeed feasible.

publication

hidden

Footer

FH Zentralschweiz

Links zu den Social-Media-Kanälen

  •  Facebook
  •  Instagram
  •  Twitter
  •  LinkedIn
  •  YouTube
  •  Flickr

Kontakt

Logo Technik & Architektur

Hochschule Luzern

Technik & Architektur

Technikumstrasse 21
CH- 6048 Horw

+41 41 349 33 11

technik-architektur@hslu.ch

Direkteinstieg

  • Für Studierende
  • Weiterbildungsinteressierte
  • Für Mitarbeitende
  • Medienschaffende

Quicklink

  • Personensuche
  • Jobs & Karriere
  • Organisation des Departements Technik & Architektur
  • Facts & Figures
  • Diversity
  • Räume mieten
  • Bibliothek

Statische Links

  • Newsletter abonnieren
  • Datenschutzerklärung
  • Impressum
  • Institutionell akkreditiert nach HFKG 2019–2026
Logo Swissuniversities

QrCode

QrCode
Wir verwenden Cookies, um Ihnen eine optimale Nutzung der Website zu ermöglichen und um Ihnen auf unserer Website, auf anderen Websites und in sozialen Netzwerken personalisierte Werbung anzuzeigen. Indem Sie diesen Hinweis schliessen oder mit dem Besuch der Seite fortfahren, akzeptieren Sie die Verwendung von Cookies. Weitere Informationen zu diesen Cookies und wie Sie die Datenbearbeitung durch sie ablehnen können, finden Sie in unserer Datenschutzerklärung.
OK