Correspondence identification for collaborative multi-robot perception under uncertainty
Correspondence identification for collaborative multi-robot perception under uncertainty
复制标题
DOI：
10.1007/s10514-021-10009-6
复制
发表时间：
2021-08
影响因子：
3.5
通讯作者：
Peng Gao;Rui Guo;Hongsheng Lu;Hao Zhang
中科院分区：
文献类型：
--
作者：
Peng Gao;Rui Guo;Hongsheng Lu;Hao Zhang
对应性识别是多机器人协作感知的关键能力，它允许一组机器人一致地指代各自视野中的相同对象。对应识别是具有挑战性的，由于存在的非共同可见的对象，不能观察到的所有机器人，并由于在机器人感知的不确定性。在本文中，我们介绍了一种新的原则性的方法，制定对应识别的正则化约束优化的数学框架下的图匹配问题。我们开发了一个正则化项，通过惩罚具有高不确定性的对象对应来明确地解决感知不确定性。我们还引入了第二个正则化项，通过惩罚非共视对象建立的对应来显式地解决非共视对象。我们的方法在机器人模拟和真实的物理机器人中进行了评估。实验结果表明，该方法能够解决不确定性和非共视性下的对应关系识别问题，达到了最佳性能。
Correspondence identification is a critical capability for multi-robot collaborative perception, which allows a group of robots to consistently refer to the same objects in their own fields of view. Correspondence identification is challenging due to the existence of non-covisible objects that cannot be observed by all robots, and due to uncertainty in robot perception. In this paper, we introduce a novel principled approach that formulates correspondence identification as a graph matching problem under the mathematical framework of regularized constrained optimization. We develop a regularization term to explicitly address perception uncertainties by penalizing the object correspondences with a high uncertainty. We also introduce a second regularization term to explicitly address non-covisible objects by penalizing the correspondences built by the non-covisible objects. Our approach is evaluated in robotic simulations and real physical robots. Experimental results show that our method is able to address correspondence identification under uncertainty and non-covisibility, and achieves the state-of-the-art performance.