Kurami, Asuka (Waseda University), Wasa, Yasuaki (Waseda University), Kishida, Masako (National Institute of Informatics)

Vision-Based Mobile Robot Control in a Floor Projected Mixed-Reality

Scheduled for presentation during the Regular Session "Advanced Motion Planning and Navigation" (FrBT2), Friday, July 18, 2025, 15:00−15:20, Room 106

Joint 10th IFAC Symposium on Mechatronic Systems and 14th Symposium on Robotics, July 15-18, 2025, Paris, France

This information is tentative and subject to change. Compiled on August 2, 2025

Abstract

This paper presents a practical approach for mobile robot control using only onboard 2D camera images and odometry data, aiming to achieve reliable navigation with minimal sensor requirements. Our system integrates two recent computer vision techniques, YOLO11 and Depth Anything V2, for robust environment recognition and safe region detection. To handle uncertainties in real-world conditions, we propose a human-in-the-loop control scheme that enables adaptive path selection in dynamic environments. Additionally, we develop a low-cost mixed reality environment that not only validates our control approach but also provides a reusable testbed for vision-based navigation systems, addressing practical challenges such as sensor drift that are often overlooked in conventional simulation environments.