Multicore bundle adjustment

TLDR

The authors design and implement inexact Newton bundle‑adjustment algorithms that leverage hardware parallelism to efficiently solve large‑scale 3D scene reconstruction. They employ multicore CPUs and GPUs to accelerate the bundle‑adjustment process. The approach yields space‑efficient algorithms and significant runtime savings, achieving up to ten‑fold speedups on CPUs and thirty‑fold on GPUs compared to state‑of‑the‑art methods while preserving convergence. Code and additional results are available at http://grail.cs.washington.edu/projects/mcba.

Abstract

We present the design and implementation of new inexact Newton type Bundle Adjustment algorithms that exploit hardware parallelism for efficiently solving large scale 3D scene reconstruction problems. We explore the use of multicore CPU as well as multicore GPUs for this purpose. We show that overcoming the severe memory and bandwidth limitations of current generation GPUs not only leads to more space efficient algorithms, but also to surprising savings in runtime. Our CPU based system is up to ten times and our GPU based system is up to thirty times faster than the current state of the art methods, while maintaining comparable convergence behavior. The code and additional results are available at http://grail.cs.washington.edu/projects/mcba.