Loading...

Development of a Vision-Based Position and Attitude Estimation Algorithm for Close Proximity Operation of Small Satellite

Pasand, Milad | 2012

641 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 43859 (45)
  4. University: Sharif University of Technology
  5. Department: Aerospace Engineering
  6. Advisor(s): Saghafi, Fariborz
  7. Abstract:
  8. One of the key technologies in the future space missions is the formation flight of small satellites. Currently, widespread research in this field is being carried out and remarkable developments have been achieved. In the mentioned laboratory researches, satellite navigation is performed by conventional sensors like infrared and ultrasonic. Both of these sensors are of active type and energy consuming. Also, use of the ultrasonic sensors is impractical in space missions (due to vacuum condition). Therefore, as an alternative for space missions, visual sensors are good candidates. The main goal of this research is to present a method for the relative navigation in proximity operations in an unknown environment using cameras installed on satellites. In this thesis, the natural feature of satellites extracted via SIFT blob detector as an affine transform invariant feature without using any artificial landmark, and the relative position and attitude is estimated by POSIT and paraperspective algorithms. Comparison of the artificial features and natural features results that the artificial features are easier and faster to extract but are limited by the number. In the other hand, if sufficient numbers of features are not visible, navigation calculations will not be possible. Therefore, in this research we use the natural features instead of the artificial ones. To simulate the proposed method, SPHERES satellite specifications are used in order to be modeled in the virtual environment
  9. Keywords:
  10. Small Satellite ; Attitude Estimation ; Natural Marks ; POSIT Algorithm ; Paraperspective

 Digital Object List

 Bookmark

No TOC