Euclidean Structure from Uncalibrated Images

Abstract

A number of recent papers have demonstrated that camera "selfcalibration" can be accomplished purely from image measurements, without requiring special calibration objects or known camera motion. We describe a method, based on self-calibration, for obtaining (scaled) Euclidean structure from multiple uncalibrated perspective images using only point matches between views.The method is in two stages. First, using an uncalibrated camera, structure is recovered up to an affine ambiguity from two views. Second, from one or more further views of this affine structure the camera intrinsic parameters are determined, and the structure ambiguity reduced to scaled Euclidean. The technique is independent of how the affine structure is obtained. We analyse its limitations and degeneracies.Results are given for images of real scenes. An application is described for active vision, where a Euclidean reconstruction is obtained during normal operation with an initially uncalibrated camera. Finally, it is demonstrated that Euclidean reconstruction can be obtained from a single perspective image of a repeated structure.

3d Modelling With Linear Approaches Using Geometric Primitives

“…Secondly, the reduced canonical parallelepiped projection matrices encode the affine properties of the scene. In consequence, the calibration problem is reduced to a self-calibration problem where the plane at infinity is already localized [3,4] or where the cameras are stationary [5][6][7].…”

“…Our calibration approach is conceptually close to selfcalibration methods, especially those that upgrade affine to euclidean structure [3,4] or those that consider special camera motions [5][6][7]. The way metric information on a parallelepiped is used is also similar to vanishing point based methods [8][9][10][11].…”

Image-Based 3D Modeling: Modeling from Reality

“…As a result, also existing footage can be used to reconstruct scenes that no longer exist. Much along the lines of work reported by Armstrong et al [1], the method is based on the automatic tracking of image features over the different views. This is done in stages.…”

Depth distortion under calibration uncertainty

“…One way to circumvent this problem is to enforce special camera displacements to obtain better estimate of the intrinsic parameters [1,7,13,29,37]. For example, a pure rotating camera can lead to a more accurate calibration parameters estimation [13].…”

“…Several major features of the resulting distortion are then made explicit. The main goals of Section 3 are: (1) to elaborate the relations between the depth distortion and the estimation of both the intrinsic and extrinsic parameters and (2) to study certain well-behaved depth distortion resulting from ambiguous solutions. Section 4 presents experiments to support the use of the visibility constraint in obtaining partial solutions to the estimation of both motion and structure.…”