Home Research Teaching Personal

  Main   Mathematical Models in Computer Vision: The Handbook, Springer (2005)

  • 3D from Image Sequences: Calibration, Motion and Shape Recovery, M. Pollefeys
ABSTRACT: In this chapter we discuss how to recover the motion and calibration of a camera and the shape of a static object from an image sequence.  The problem can be split into four subproblems: (1) computing the geometric relation between neighboring images, (2) estimating the motion and calibration of the camera, (3) computing a dense set of correspondences between neighboring images, (4) reconstruction of the 3D object shape.  The approach we present here is fully automatic and can deal with photo or video sequences acquired with an uncalibrated hand-held camera.  The different algorithms can also be used to provide solutions for other applications. 

Last Update: December20th, 2004, you can mail your comments to: nikos.paragios@computer.org