Vision based Autonomous Navigation and Human Robot Interaction (HRI) for a Team of Flying Robots

飞行机器人团队基于视觉的自主导航和人机交互 (HRI)

• 批准号: 472804-2015
• 负责人: Tan, Ping
• 金额: $ 8.23万
• 依托单位: Simon Fraser University
• 依托单位国家: 加拿大
• 项目类别: Research Tools and Instruments - Category 1 (<$150,000)
• 财政年份: 2014
• 资助国家: 加拿大
• 起止时间: 2014-01-01 至 2015-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=564849
• 关键词: Vision; based; Autonomous; Navigation; Human

We propose to purchase a high-quality indoor motion-capture system to track the locations of a team of flying robots and humans as they interact, cameras for data collection by a team of robots, and computer servers and storage to process these data. This equipment will support the applicants' individual and joint research programs. Each of our team has recently demonstrated world-leading new capabilities in multi-camera tracking and SLAM (Tan), in vision-mediated human-robot interaction (Vaughan), and human activity recognition from video (Mori). The proposed equipment will support our new joint work, where we (i) control a team of flying robots to observe moving people from several viewpoints (ii) reconstruct a 3D view of each person and extract a skeleton pose estimate in real time, (iii) categorize the people's activity, and (iv) use the activity information to help the robots improve their camera placement.

我们建议购买一个高质量的室内运动捕捉系统，以跟踪一组飞行机器人和人类互动时的位置，一组机器人收集数据的摄像头，以及处理这些数据的计算机服务器和存储器。这些设备将支持申请人的个人和联合研究计划。我们的每个团队最近都展示了世界领先的多摄像头跟踪和SLAM（Tan），视觉介导的人机交互（Vaughan）和视频人类活动识别（Mori）的新功能。拟议的设备将支持我们的新的联合工作，我们（i）控制一组飞行机器人从几个角度观察移动的人（ii）重建每个人的3D视图，并提取骨架姿态估计在真实的时间，（iii）分类的人的活动，以及（iv）使用活动信息，以帮助机器人改善他们的相机位置。