Visual Odometry

Basic Concepts:

  • It should be possible to determine the 3D rotation and translation of the robot from optical flow vectors.
  • Optical flow vectors can be generated from the coordinates of corresponding points in successive images.
  • The OpenCV Computer Vision Library includes routines to automatically identify optical flow vectors from image sequences:  http://www.intel.com/technology/computing/opencv/index.htm
  • It should be possible to use landmark recognition to correct for cumulative position and orientation errors that otherwise could have a catastrophic impact on the absolute cumulative error.

Relevant Publications:

Omnidirectional Visual Odometry for a Planetary Rover

 Peter Corke, Dennis Strelow, and Sanjiv Singh

Carnegie Mellon University

http://www.ri.cmu.edu/pub_files/pub4/corke_peter_ian_2004_1/corke_peter_ian_2004_1.pdf

 

An Empirical Comparison of Methods for Image-Based Motion Estimation

Henele Adams, Sanjiv Singh, and Dennis Strelow

Carnegie Mellon University

http://www.ri.cmu.edu/pub_files/pub4/adams_henele_2002_1/adams_henele_2002_1.pdf

 

Visual Odometry Using Commodity Optical Flow

Jason Campbell, Rahul Sukthankar, and Illah Nourbakhsh

Intel, Carnegie Mellon University

http://info.pittsburgh.intel-research.net/People/jasonc/pubs/Campbell_VisOdom_AAAI2004_Demo_Overview.pdf

 

Techniques for Evaluating Optical Flow

for Visual Odometry in Extreme Terrain

Jason Campbell, Rahul Sukthankar, and Illah Nourbakhsh

Intel, Carnegie Mellon University, NASA

http://info.pittsburgh.intel-research.net/People/jasonc/pubs/Campbell_VisOdomMeasurement_IROS2004_Paper.pdf

 

Visual Odometry and Map Correlation

Anat Levin and Richard Szeliski

Microsoft

http://research.microsoft.com/vision/visionbasedmodeling/publications/Levin-CVPR04.pdf

 

Visual Odometry

David Nist´er, Oleg Naroditsky, and James Bergen

Sarnoff Corporation

http://www.vis.uky.edu/~dnister/Publications/2004/VisualOdometry/nister_d_146.pdf

 

A Robust Visual Odometry and Precipice Detection System Using Consumer-grade Monocular Vision

Jason Campbell, Rahul Sukthankar, Illah Nourbakhsh, and Aroon Pahwa

Intel, Carnegie Mellon University NASA

http://www.cs.cmu.edu/~personalrover/PER/ResearchersPapers/CampbellSukthankarNourbakhshPahwa_VisualOdometryCR.pdf

   

Stereo-Based Ego-Motion Estimation

Using Pixel Tracking and Iterative Closest Point

 Annalisa Milella and Roland Siegwart

Politecnico of Bari, Swiss Federal Institute of Technology

http://asl.epfl.ch/aslInternalWeb/ASL/publications/uploadedFiles/amilella_EgoMotion.rev.pdf