Robust Stereo Visual SLAM for Dynamic Environments With Moving Object

Gang Li,Xiang Liao,Huilan Huang,Shaojian Song,Bin Liu,Yawen Zeng

DOI: https://doi.org/10.1109/access.2021.3059866

IF: 3.9

2021-01-01

IEEE Access

Abstract:The accuracy of localization and mapping of automated guided vehicles (AGVs) using visual simultaneous localization and mapping (SLAM) is significantly reduced in a dynamic environment compared to a static environment due to incorrect data association caused by dynamic objects. To solve this problem, a robust stereo SLAM algorithm based on dynamic region rejection is proposed. The algorithm first detects dynamic feature points from the fundamental matrix of consecutive frames and then divides the current frame into superpixels and labels its boundaries with disparity. Finally, dynamic regions are obtained from dynamic feature points and superpixel boundaries types; only the static area is used to estimate the pose to improve the localization accuracy and robustness of the algorithm. Experiments show that the proposed algorithm outperforms ORB-SLAM2 in the KITTI dataset, and the absolute trajectory error in the actual dynamic environment can be reduced by 84% compared with the conventional ORB-SLAM2, which can effectively improve the localization and mapping accuracy of AGVs in dynamic environments.

computer science, information systems,telecommunications,engineering, electrical & electronic

What problem does this paper attempt to address?

### Main Problems Addressed by the Paper

This paper primarily focuses on the research of Visual Simultaneous Localization and Mapping (VSLAM) technology for Automated Guided Vehicles (AGVs) in dynamic environments. Specifically, the paper aims to address the following core issues:

1. **Impact of Dynamic Objects on Localization Accuracy**: In dynamic environments, the presence of dynamic objects leads to data association errors, significantly reducing the accuracy of AGVs' localization and mapping using VSLAM technology.

2. **Proposing a Robust Stereo Vision SLAM Algorithm**: To tackle the aforementioned challenges, the paper proposes a robust stereo vision SLAM algorithm based on dynamic region removal. This algorithm first detects dynamic feature points and segments the current frame into superpixels, then labels the boundary types. Finally, it obtains dynamic regions through dynamic feature points and superpixel boundary types, and uses only static regions to estimate the pose, thereby improving localization accuracy and robustness.

3. **Experimental Validation**: Experiments on the KITTI dataset and in real dynamic environments show that the proposed algorithm reduces the absolute trajectory error by 84% compared to ORB-SLAM2, effectively enhancing the localization and mapping accuracy of AGVs in dynamic environments.

### Main Contributions

1. **Stereo SLAM Algorithm with Dynamic Region Removal**: The paper proposes a stereo SLAM algorithm based on dynamic region removal, which includes a novel dynamic region detection algorithm. This algorithm can effectively identify and remove dynamic regions independently of localization accuracy impact.

2. **Improved SLIC Algorithm**: To eliminate redundant computations and improve the speed of dynamic region detection, the paper improves the SLIC algorithm.

3. **Extensive Experimental Validation**: The effectiveness of the proposed SLAM algorithm is validated through experiments on the KITTI dataset and in real scenarios, particularly demonstrating its performance in various dynamic environments.