TOSS: Real-time Tracking and Moving Object Segmentation for Static Scene Mapping

Seoyeon Jang,Minho Oh,Byeongho Yu,I Made Aswin Nahrendra,Seungjae Lee,Hyungtae Lim,Hyun Myung

2024-08-10

Abstract:Safe navigation with simultaneous localization and mapping (SLAM) for autonomous robots is crucial in challenging environments. To achieve this goal, detecting moving objects in the surroundings and building a static map are essential. However, existing moving object segmentation methods have been developed separately for each field, making it challenging to perform real-time navigation and precise static map building simultaneously. In this paper, we propose an integrated real-time framework that combines online tracking-based moving object segmentation with static map building. For safe navigation, we introduce a computationally efficient hierarchical association cost matrix to enable real-time moving object segmentation. In the context of precise static mapping, we present a voting-based method, DS-Voting, designed to achieve accurate dynamic object removal and static object recovery by emphasizing their spatio-temporal differences. We evaluate our proposed method quantitatively and qualitatively in the SemanticKITTI dataset and real-world challenging environments. The results demonstrate that dynamic objects can be clearly distinguished and incorporated into static map construction, even in stairs, steep hills, and dense vegetation.

Robotics

What problem does this paper attempt to address?

The paper aims to address the issues of real-time navigation and static scene mapping for autonomous robots in complex environments, particularly focusing on the challenges between moving object detection and static map construction. Specifically, the goals of the paper can be summarized as follows:

1. **Integration of Moving Object Segmentation and Static Map Construction**: Existing methods often handle moving object segmentation (MOS) and static map construction separately, making it difficult to achieve precise map construction during navigation. Therefore, the paper proposes an integrated framework (TOSS) that enables real-time moving object tracking and static map construction.

2. **Efficient Object Tracking**: To achieve real-time navigation, the paper introduces a computationally efficient hierarchical association cost matrix to quickly determine the position changes of moving objects, thereby reducing the time complexity of object tracking from the traditional O(N^2) to O(N).

3. **Improving Map Quality**: For the removal of dynamic objects and the restoration of static objects, the paper proposes a voting-based method (DS-Voting) that reduces misjudgments by emphasizing the spatial and temporal differences between dynamic and static objects, thereby improving the quality of the static map.

In summary, the core objective of the paper is to develop a system capable of real-time moving object detection and static map construction in complex environments to support the safe navigation and high-quality map construction of autonomous robots. This approach not only considers the detection of moving objects but also focuses on effectively removing the influence of these moving objects on the static map to generate a more accurate environmental model.