arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.01797cs.ROcs.CVcs.LG
收藏

利用实时的机器人视觉实现与物体无关的人-机器人交接

Object-Independent Human-to-Robot Handovers using Real Time Robotic Vision

TU Wien(维也纳工业大学);Australian Center of Robotic Vision(澳大利亚机器人视觉中心);Monash University(蒙纳士大学);University of Birmingham(伯明翰大学);Queensland University of Technology(昆士兰科技大学)

Patrick Rosenberger, Akansel Cosgun, Rhys Newbury, Jun Kwan, Valerio Ortenzi, Peter Corke, Manfred Grafinger

摘要:

我们提出了一种方法,安全和对象无关的人对机器人的实时的机器人视觉和操纵。我们的目标是通用对象检测器的普遍适用性,快速抓取选择算法,并通过使用单个夹具安装的RGB-D相机,因此不依赖于外部传感器。机器人通过视觉伺服被控制朝向感兴趣的对象。高度重视安全性,我们使用两个感知模块:人体部位分割和手/手指分割。被认为属于人类的像素从候选抓取姿势中过滤掉,从而确保机器人安全地拾取物体而不会与人类伙伴发生碰撞。抓取选择和感知模块实时并行运行,这允许监控进度。在13个物体的实验中,机器人能够在81.9%的试验中成功地从人类手中拿走物体。

英文摘要:

We present an approach for safe and object-independent human-to-robot handovers using real time robotic vision and manipulation. We aim for general applicability with a generic object detector, a fast grasp selection algorithm and by using a single gripper-mounted RGB-D camera, hence not relying on external sensors. The robot is controlled via visual servoing towards the object of interest. Putting a high emphasis on safety, we use two perception modules: human body part segmentation and hand/finger segmentation. Pixels that are deemed to belong to the human are filtered out from candidate grasp poses, hence ensuring that the robot safely picks the object without colliding with the human partner. The grasp selection and perception modules run concurrently in real-time, which allows monitoring of the progress. In experiments with 13 objects, the robot was able to successfully take the object from the human in 81.9% of the trials.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑