Bundle Ajustment -- A Modern Synthesis

Bill Triggs (1) , Philip Mclauchlan (2) , Richard Hartley (3) , Andrew Fitzgibbon (4)

1 MOVI - Modeling, localization, recognition and interpretation in computer vision
2 EE - Department of Electronic Engineering
3 General Electric [Schenectady]
4 VGG - Visual Geometry Group

This paper is a survey of the theory and methods of photogrammetric bundle adjustment, aimed at potential implementors in the computer vision community. Bundle adjustment is the problem of refining a visual reconstruction to produce jointly optimal structure and viewing parameter estimates. Topics covered include: the choice of cost function and robustness; numerical optimization including sparse Newton methods, linearly convergent approximations, updating and recursive methods; gauge (datum) invariance; and quality control. The theory is developed for general robust cost functions rather than restricting attention to traditional nonlinear least squares.

Mots clés: Gauge Freedom; Sparse Matrices; Optimization; Bundle Adjustment; Scene Reconstruction

Domaines: Synthèse d'image et réalité virtuelle [cs.GR]

Bill Triggs, Philip Mclauchlan, Richard Hartley, Andrew Fitzgibbon. Bundle Ajustment -- A Modern Synthesis. Bill Triggs and Andrew Zisserman and Richard Szeliski. Vision Algorithms: Theory and Practice, 1883, Springer-Verlag, pp.298--372, 2000, Lecture Notes in Computer Science (LNCS), ⟨10.1007/3-540-44480-7_21⟩. ⟨inria-00590128⟩