Real-time Estimation of Camera Motion and Other Endeavors in Structure from Motion
Time and Place
Auditorium (NSH 1305)
A system that estimates the motion of a single moving camera or stereo head based on video input will be presented. The system operates in real-time with low delay. The front end of the system is a feature tracker. Point features are matched between pairs of frames and linked into image trajectories at video rate. Robust estimates of the camera motion are then produced from the feature tracks using a geometric hypothesize-and-test architecture. No prior knowledge of the scene or the camera motion is necessary. Applications include navigation and robotics, augmented reality and scene reconstruction. The image-based information can also be used in conjunction with information from other sources such as GPS, inertia sensors, wheel encoders, etc. The pose estimation method has been applied successfully to video from aerial, automotive and handheld platforms.
Results with an autonomous ground vehicle will be shown, including examples of camera trajectories estimated purely from images over previously unseen distances and periods of time. The system will be shown processing real-time, and applications to scene reconstruction discussed.
David Nister received the
PhD degree in computer vision, numerical analysis and computing science from
the Royal Institute of Technology (KTH),
For appointments, please contact Janice Brochetti (email@example.com ).