Navigating a Service Robot for Indoor Complex Environments
Navigating a Service Robot for Indoor Complex Environments
复制标题
DOI：
10.3390/app9030491
复制
发表时间：
2019-02-01
影响因子：
2.7
通讯作者：
Lee, Jiann-Der
中科院分区：
文献类型：
--
作者：
Chien, Jong-Chih;Dang, Zih-Yang;Lee, Jiann-Der
本文研究了使用自主服务机器人在室内复杂的环境，如医院病房或养老院。这种类型的服务机器人不仅需要规划和寻找障碍物周围的路径，还必须与护理人员或患者进行互动。本研究提出了一种类型的服务机器人，它结合了从一个3D深度摄像头与红外传感器的图像，并从多个声纳传感器的自适应神经模糊推理系统（ANFIS）为基础的方法在路径规划的输入。在个人接触中，面部特征用于执行人员识别，以区分工作人员、患者或陌生人。在工作人员的情况下，服务机器人可以根据要求执行跟随我的功能。机器人还可以使用一个额外的功能，即对人的性别进行分类。面部和性别识别的目的包括帮助用户选择合适的目的地。实验是在狭窄但开放的空间以及狭窄的通道场景中进行的，几乎在所有情况下，自主机器人都能够到达目的地。
This paper investigates the use of an autonomous service robot in an indoor complex environment, such as a hospital ward or a retirement home. This type of service robot not only needs to plan and find paths around obstacles, but must also interact with caregivers or patients. This study presents a type of service robot that combines the image from a 3D depth camera with infrared sensors, and the inputs from multiple sonar sensors in an Adaptive Neuro-Fuzzy Inference System (ANFIS)-based approach in path planning. In personal contacts, facial features are used to perform person recognition in order to discriminate between staff, patients, or a stranger. In the case of staff, the service robot can perform a follow-me function if requested. The robot can also use an additional feature which is to classify the person's gender. The purpose of facial and gender recognition includes helping to present choices for suitable destinations to the user. Experiments were done in cramped but open spaces, as well as confined passages scenarios, and in almost all cases, the autonomous robots were able to reach their destinations.