Multigrid for Bundle Adjustment

TL;DR

This paper introduces a multigrid preconditioner for bundle adjustment that significantly improves the speed of solving large-scale problems, achieving up to 13 times faster performance than current methods.

Contribution

It proposes an unsmoothed aggregation multigrid preconditioner that better captures global modes, addressing superlinear scaling issues in large bundle adjustment problems.

Findings

Achieves up to 13x faster solutions on large problems.

Addresses superlinear scaling in existing bundle adjustment methods.

Demonstrates improved conditioning with multigrid preconditioning.

Abstract

Bundle adjustment is an important global optimization step in many structure from motion pipelines. Performance is dependent on the speed of the linear solver used to compute steps towards the optimum. For large problems, the current state of the art scales superlinearly with the number of cameras in the problem. We investigate the conditioning of global bundle adjustment problems as the number of images increases in different regimes and fundamental consequences in terms of superlinear scaling of the current state of the art methods. We present an unsmoothed aggregation multigrid preconditioner that accurately represents the global modes that underlie poor scaling of existing methods and demonstrate solves of up to 13 times faster than the state of the art on large, challenging problem sets.