CodedVO: Coded Visual Odometry

TL;DR

CodedVO introduces a novel monocular visual odometry approach that uses custom optics to encode metric depth, overcoming scale ambiguity and achieving high accuracy in indoor environments.

Contribution

The paper presents CodedVO, a new method that physically encodes depth into images to improve monocular odometry accuracy.

Findings

Achieves 0.08m average trajectory error on ICL-NUIM dataset.

Demonstrates robustness and adaptability in diverse indoor environments.

Overcomes scale ambiguity in monocular visual odometry.

Abstract

Autonomous robots often rely on monocular cameras for odometry estimation and navigation. However, the scale ambiguity problem presents a critical barrier to effective monocular visual odometry. In this paper, we present CodedVO, a novel monocular visual odometry method that overcomes the scale ambiguity problem by employing custom optics to physically encode metric depth information into imagery. By incorporating this information into our odometry pipeline, we achieve state-of-the-art performance in monocular visual odometry with a known scale. We evaluate our method in diverse indoor environments and demonstrate its robustness and adaptability. We achieve a 0.08m average trajectory error in odometry evaluation on the ICL-NUIM indoor odometry dataset.