.png){kind=icon}
Return
HUMAN FOLLOWING TASK BASED ON ALGORITHM
.png){kind=icon}
PRE
AI
DOI：10.2316/J.2026.206-1237.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In recent years, the task of human following for mobile robots has received considerable attention in the field of human-computer interaction. However, the perception strategies currently employed by mobile robots lead to the loss of the target person, and the widely applied path planning algorithms are not performing well in terms of real-time performance and versatility. For these issues, we propose a multi-sensor fusion-based identification and localisation of the target person (ILTP) method, as well as an Adaptive parameter dynamic window algorithm (APDWA). Firstly, we fuse visual sensors with gimbals, LiDAR and ultra-wide band (UWB) sensors to address the problem of target person loss during the person-following process of mobile robots, which is caused by limited onboard field of view (FOV) and complex environments. Meanwhile, we propose an adaptive parameter strategy based on the state of the following target to reduce redundant predicted paths. Finally, the proposed ILTP method is compared with the currently widely used single-sensor perception strategy, and the APDWA is compared with metaheuristic, deep learning-based and classical algorithms, verifying its rapidity and robustness in the human following task.
Keywords:
Human following
mobile robots
target localisation
path planning
human-robot collaboration
Journal
I
IF:
1
Papers:
34
Citations:
0
Organization
Cited Papers
No cited papers available