| Title |
Map Merging Using Temporal and Probabilistic Reliability Maps in Environments with Dynamic Objects |
| Authors |
김한규(Hann-Gyoo Kim) ; 이승환(Seung-Hwan Lee) |
| DOI |
https://doi.org/10.5370/KIEE.2025.74.4.677 |
| Keywords |
Map Matching; Map Merging; Temporal Map; Probabilistic Reliability Map; Dynamic Objects; Multi-robot System |
| Abstract |
This paper proposes a novel map merging technique designed to improve accuracy and reliability in environments with multiple dynamic objects by using temporal maps and probabilistic reliability maps. Traditional map merging methods often experience reduced accuracy in such settings due to the significant spatial and temporal variability introduced by dynamic objects. The proposed method addresses this by recording object update times in a temporal map and evaluating object reliability over time with a probabilistic reliability map, enabling effective handling of uncertainties caused by dynamic objects. In particular, the probabilistic reliability map estimates object reliability over time, allowing for the selective exclusion of less reliable data and ensuring a more precise merged map. Simulation results demonstrate that the proposed method outperforms existing approaches across various environments and scenarios, both visually and quantitatively. Future research will focus on extending this technique to 3D environments and validating its performance in real-world settings to enhance its practical applicability. |