Inria Grenoble - Rhône-Alpes, UJF - Université Joseph Fourier - Grenoble 1, INPG - Institut National Polytechnique de Grenoble , CNRS - Centre National de la Recherche Scientifique : UMR5217
Abstract : We present a new strategy for RANSAC sampling named BetaSAC, in reference to the beta distribution. Our proposed sampler builds a hypothesis set incrementally, select- ing data points conditional on the previous data selected for the set. Such a sampling is shown to provide more suitable samples in terms of inlier ratio but also of consistency and potential to lead to an accurate parameters estimation. The algorithm is presented as a general framework, easily implemented and able to exploit any kind of prior infor- mation on the potential of a sample. As with PROSAC, BetaSAC converges towards RANSAC in the worst case. The benefits of the method are demonstrated on the homog- raphy estimation problem.
https://hal.inria.fr/hal-00669125
Contributor : James Crowley
<>
Submitted on : Saturday, February 11, 2012 - 3:06:10 PM
Last modification on : Thursday, February 7, 2019 - 3:48:00 PM
Long-term archiving on : Thursday, November 22, 2012 - 12:05:30 PM
Antoine Meler, Marion Decrouez, James L. Crowley. BetaSAC: A New Conditional Sampling For RANSAC. British Machine Vision Conference 2010, British Machine Vision Association, Aug 2010, Aberystwyth, United Kingdom. ⟨hal-00669125⟩
