Particle filter approach adapted to catadioptric images for target tracking application

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 59
  • Download : 0
This paper addresses the problem of object tracking by particle filter in catadioptric images. Whereas a large literature exists for traditional images, only a very few methods have been developed for catadioptric vision. We present two techniques to correctly deal with the strong distortion inherent to catadioptric images. First, we propose generating and diffusing the particles in the equivalent sphere space rather than in the original 2D catadioptric image. It allows to handle the specific distance associated to the distorted image in a general framework. Second, we use an adapted neighborhood to perform template matching by histogram comparison. It permits to compare templates quickly while using an active neighborhood for space-variant windows. An additional important feature of our system is that no interpolation is performed: we directly work on the original catadioptric image. Experimental results demonstrated the validity of the proposed approach. © 2009. The copyright of this document resides with its authors.
Publisher
British Machine Vision Association, BMVA
Issue Date
2009-09-07
Language
English
Citation

2009 20th British Machine Vision Conference, BMVC 2009

DOI
10.5244/C.23.37
URI
http://hdl.handle.net/10203/225788
Appears in Collection
GCT-Conference Papers(학술회의논문)
Files in This Item
There are no files associated with this item.

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0