Sciweavers

RAS
2006

Efficient visual homing based on Fourier transformed panoramic images

13 years 2 months ago
Efficient visual homing based on Fourier transformed panoramic images
We present a fast and efficient homing algorithm based on Fourier transformed panoramic images. By continuously comparing Fourier coefficients calculated from the current view with coefficients representing the goal location, a mobile robot is able to find its way back to known locations. No prior knowledge about the orientation with respect to the goal location is required, since the Fourier phase is used for a fast sub-pixel orientation estimation. We present homing runs performed by an autonomous mobile robot in an office environment. In a more comprehensive investigation the algorithm is tested on an image data base recorded by a small mobile robot in a toy house arena. Catchment areas for the proposed algorithm are calculated and compared to results of a homing scheme described in [M. Franz, B. Sch
Wolfgang Stürzl, Hanspeter A. Mallot
Added 14 Dec 2010
Updated 14 Dec 2010
Type Journal
Year 2006
Where RAS
Authors Wolfgang Stürzl, Hanspeter A. Mallot
Comments (0)