Maximum Likelihood Combining of Stochastic Maps

Abstract

The problem of combining stochastic maps obtained by independent agents is considered. Using generalized likelihood ratio statistics, the problem of matching triangles that correspond to common landmark observations in different stochastic maps is formulated as a bipartite matching problem with generalized likelihood ratio statistics. From the matched triangles between each map, the maximum likelihood combining of stochastic maps is generated. It is shown that the generalized likelihood ratio statistic and the maximum likelihood combining of maps can be computed in closed form, which makes the proposed algorithm a scalable solution to matching and combining stochastic maps with a large number of landmarks.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2011
Accession Number
ADA557935

Entities

People

  • Brandon A. Jones
  • Lang Tong
  • Mark Campbell

Organizations

  • Cornell University

Tags

Communities of Interest

  • Autonomy
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Algorithms
  • Autonomous Navigation
  • Computations
  • Coordinate Systems
  • Data Association
  • Detection
  • False Alarms
  • Gaussian Noise
  • Linear Programming
  • Optimization
  • Pattern Recognition
  • Robot Navigation
  • Robots
  • Simultaneous Localization And Mapping
  • Statistics
  • Triangulation
  • World Geodetic System

Readers

  • Computer Vision.
  • Graph Algorithms and Convex Optimization.
  • Statistical inference.