Difference between revisions of "Projects:PointSetRigidRegistration"

From NAMIC Wiki
Jump to: navigation, search
Line 3: Line 3:
  
 
= Particle Filtering with Stochastic Dynamics for Point Set Registration =
 
= Particle Filtering with Stochastic Dynamics for Point Set Registration =
Typically, registration algorithms compute the transformations paramaters by maximizing a metric given an estimate of the correspondence between points across the two sets of interest.  This can be viewed as a posterior estimation problem problem, in which the corresponding distribution can naturally be estimated using a particle filter.  Moreover, we treat motion as a local variation in the pose parameters obatined from running a few iterations of the standard Iterative Closest Point (ICP) algorithm.  Employing this idea, we introduce stochastic motion dynamics to widen the narrow band of convergence as well as provide a dynamical model of uncertainity.
+
Typically, registration algorithms compute the transformations paramaters by maximizing a metric given an estimate of the correspondence between points across the two sets of interest.  This can be viewed as a posterior estimation problem problem, in which the corresponding distribution can naturally be estimated using a particle filter.  In this work, we treat motion as a local variation in the pose parameters obatined from running a few iterations of the standard Iterative Closest Point (ICP) algorithm.  Employing this idea, we introduce stochastic motion dynamics to widen the narrow band of convergence as well as provide a dynamical model of uncertainity.  In contrast with other techniques, our approach requires no annealing schedule, which results in a reduction in computational complexity as well as maintains the temoral coherency of the state (no loss of information).  Also, unlike most alternative approaches for point set registration, we make no geometric assumptions on the two data sets.
 
= Description =
 
= Description =
  

Revision as of 02:50, 27 April 2008

Home < Projects:PointSetRigidRegistration
Back to NA-MIC Collaborations, Georgia Tech Algorithms


Particle Filtering with Stochastic Dynamics for Point Set Registration

Typically, registration algorithms compute the transformations paramaters by maximizing a metric given an estimate of the correspondence between points across the two sets of interest. This can be viewed as a posterior estimation problem problem, in which the corresponding distribution can naturally be estimated using a particle filter. In this work, we treat motion as a local variation in the pose parameters obatined from running a few iterations of the standard Iterative Closest Point (ICP) algorithm. Employing this idea, we introduce stochastic motion dynamics to widen the narrow band of convergence as well as provide a dynamical model of uncertainity. In contrast with other techniques, our approach requires no annealing schedule, which results in a reduction in computational complexity as well as maintains the temoral coherency of the state (no loss of information). Also, unlike most alternative approaches for point set registration, we make no geometric assumptions on the two data sets.

Description

Algorithm

Project Status

Key Investigators

  • Georgia Tech: Romeil Sandhu, Samuel Dambreville, Allen Tannenbaum

Publications

In press

  • R. Sandhu, S. Dambreville, A. Tannenbaum. Particle Filtering for Registration of 2D and 3D Point Sets with Stochastic Dynamics. In CVPR, 2008.