Curve Finder Combining Perceptual Grouping and a Kalman Like Fitting

Frederic Guichard and Jean-Philippe Tarel

Abstract

We present an algorithm that extracts curves from a set of edgels within a specific class in a decreasing order of their ``length''. The algorithm inherits the perceptual grouping approaches. But, instead of using only local cues, a global constraint is imposed to each extracted subset of edgels, that the underlying curve belongs to a specific class. In order to reduce the complexity of the solution, we work with a linearly parameterized class of curves, function of one image coordinate. This allows, first, to use a recursive Kalman based fitting and, second, to cast the problem as an optimal path search in an directed graph. Experiments on finding lane-markings on roads demonstrate that real-time processing is achievable.

Reference

@INPROCEEDINGS{jpt-iccv99, 
  author    = {Guichard, F. and Tarel, J.-P.}, 
  title     = {Curve Finder Combining Perceptual Grouping and a Kalman Like Fitting},
  booktitle = {IEEE International Conference on Computer Vision (ICCV'99)}, 
  date      = {September 20-25},
  address   = {Kerkyra, Greece},
  page      = {1003-1008},
  year      = {1999},
  note      = {http://perso.lcpc.fr/tarel.jean-philippe/iccv99.html}
}

Compressed postscript file (482 Kb)
Pdf file (432 Kb)
(c) IEEE