3D Structure From Visual Motion

Course Aim & Organization

This is a description page for the PhD course on 3D Structure from Visual Motion: Novel Techniques in Computer Vision and Autonomous Robots/Vehicles. It is meant to present modern techniques to simultaneously estimate the unknown motion of a camera while reconstructing the 3D structure of the observed world to be applied un scientific fields such as: 3D reconstruction, autonomous robot navigation, aerial/field surveying, unmanned vehicle maneuvering, etc.

Teachers

Although formally entitled to just one of the teachers the course is held by (in order of appearance)

• Matteo Matteucci
• Davide Migliore
• Vincenzo Caglioti
• Marco Marcon
• Domenico G. Sorrenti

Course Schedule

This is the schedule foreseen for the course. The timing refers to the duration of the room reservation not necessarily the duration of the lecture ;-)

• 15/02/2010 14:30-18:30 in Sala Seminari DEI
  • Course introduction (M. Matteucci)
  • Correspondence analysis: tracking and ransac (D. Migliore)
  • Optical flow (D. Migliore)
• 17/02/2010 14:30-18:30 in Aula PT1 DEI
  • Projection model and projection matrix (V. Caglioti)
  • Fundamental and Essential matrices (V. Caglioti)
• 22/02/2010 14:30-18:30 in Sala Seminari DEI
  • Motion extraction and 3D reconstruction (V. Caglioti)
• 24/02/2010 14:30-18:30 in Sala Seminari DEI
  • Combined estiamation of 3D structure and camera egomotion (M.Marcon) (M. Marcon)
• 26/02/2010 14:30-18:30 in Sala Seminari DEI
  • Uncalibrated visual odometry (V. Caglioti)
  • Omnidirectional odometry (V. Caglioti)
• 01/03/2010 14:30-18:30 in Aula PT1
  • Bayesian Filtering and SLAM (M. Matteucci)
• 03/03/2010 14:30-18:30 in Aula PT1 DEI
  • MonoSLAM, PTAM and FrameSLAM (M. Matteucci, D.G. Sorrenti)
• 05/03/2010 14:30-18:30 in Sala Seminari DEI
  • 3D without 3D: plenoptic methods, lumigraph, albedo, non Lambertian surfaces (M. Marcon)

Course Material & Referencies

The following is some suggested material to follow the course lectures.

Slides and lecture notes

• Camera geometry, single view, and two view geometry material
• Two view geometry and visual odometry material
• Correspondence analysis and RANSAC
• Optical flow tracking and egomotion estimation
• Structure from Motion
• Bayesian Filtering, Kalman Filtering, and SLAM
• Simultaneous Localization and Mapping a.k.a. SLAM!
• Monocular SLAM
• Stereo and Omnidirectional SLAM
• Panoramic Visual Odomentry
• Parallel Tracking and Mapping
• 3D without 3D

Suggested Bibliography

• R. Hartley, A. Zisserman. Multiple View Geometry in Computer Vision, Cambridge University Press, March 2004.
• S. Thrun, W. Burgard, D. Fox. Probabilistic Robotics, MIT Press, September 2005.
• Papers you might find useful to deepen your study:
  • ...