ASL Student Project Description

Transcrição

ASL Student Project Description
Autonomous Systems Lab
Prof. Dr. Roland Siegwart
Satellite tracking from camera: the inverse GPS problem
The student will develop methods to track GPS satellites from mobile vehicles. The student has to
draw the expected satellite position on an image coming from a camera pointing the sky, given the
vehicle approximate position. This will be useful to estimate in urban environments the reliability of
GPS signal (sky occlusion → bad GPS estimate)
Typ / type
Partner
Semester Project
ETHZ – ASL lab
Zeitdauer / period
Student(en) /
student(s)
Betreuung intern /
internal supervisor
Betreuung extern /
external supervisor
Winter semester
Luciano Spinello ([email protected])
Dr. Cedric Predalier ([email protected])
Stichworte / key words
The student will develop methods to track GPS satellites from mobile vehicles. The student has to
draw the expected satellite position on an image coming from a camera pointing the sky, given the
vehicle approximate position. This will be useful to estimate in urban environments the reliability of
GPS signal (sky occlusion → bad GPS estimate)
If, for example, the expected satellite position in the camera coincides with a buidling in the image then
the GPS estimate is unreliable (due to multipath problems and signal bouncings) and the robot
localization can be corrected
The algorithm has to be developed in C++/Matlab using standard computer vision libraries and open
source tools.
The work will be divided as:
−
literature study (review of the current work in the field)
−
theoretical development of the method
−
C++/Matlab development