期刊
IEEE ACCESS
卷 9, 期 -, 页码 85265-85276出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/ACCESS.2021.3082947
关键词
Robots; Cameras; Data models; Robot kinematics; Training; Three-dimensional displays; Robot vision systems; Monocular vision; reinforcement learning; reward shaping; robot manipulation
资金
- Key Research Development Program (KRDP) of Anhui Province [201904a05020086]
- NSFC [61804100, 61973294, 61806181]
- Chinese Academy of Science (CAS) [GJTD-2018-15]
A calibration-free monocular vision-based robot manipulation approach using domain randomization and deep reinforcement learning (DRL) is proposed to estimate spatial information of targets in a large manipulation environment and actively avoid occlusions during robot manipulations.
Vision-based manipulation has been largely used in various robot applications. Normally, in order to obtain the spatial information of the operated target, a carefully calibrated stereo vision system is required. However, it limits the application of robots in the unstructured environment which limits both the number and the pose of the camera. In this study, a calibration-free monocular vision-based robot manipulation approach is proposed based on domain randomization and deep reinforcement learning (DRL). Firstly, a learning strategy combined domain randomization is developed to estimate the spatial information of the target from a single monocular camera arbitrarily mounted in a large area of the manipulation environment. Secondly, to address the monocular occlusion problem which regularly happens during robot manipulations, an occlusion awareness DRL policy has been designed to control the robot to avoid occlusions actively in the manipulation tasks. The performance of our method has been evaluated on two common manipulation tasks, reaching and lifting of a target building block, which show the efficiency and effectiveness of our proposed approach.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据