Dynamic Registration: Joint Ego Motion Estimation and 3D Moving Object Detection in Dynamic Environment

TL;DR

This paper introduces Dynamic Registration, a general method for simultaneous ego-motion estimation and dynamic object detection in lidar-based localization, improving accuracy in dynamic environments without relying on a new SLAM framework.

Contribution

It proposes a novel iterative approach that integrates dynamic object removal with ego-motion estimation, compatible with any lidar SLAM system, and demonstrates improved performance on KITTI datasets.

Findings

Stable and consistent improvements over classical registration algorithms.

Effective segmentation of dynamic objects and static environment.

Applicable to various lidar SLAM systems.

Abstract

Localization in a dynamic environment suffers from moving objects. Removing dynamic object is crucial in this situation but become tricky when ego-motion is coupled. In this paper, instead of proposing a new slam framework, we aim at a more general strategy for a localization scenario. In that case, Dynamic Registration is available for integrating with any lidar slam system. We utilize 3D object detection to obtain potential moving objects and remove them temporarily. Then we proposed Dynamic Registration, to iteratively estimate ego-motion and segment moving objects until no static object generates. Static objects are merged with the environment. Finally, we successfully segment dynamic objects, static environments with static objects, and ego-motion estimation in a dynamic environment. We evaluate the performance of our proposed method on KITTI Tracking datasets. Results show stable and consistent improvements based on other classical registration algorithms.