首页 /研究 /Development of Optical Communication Marks for Mobile Robots to Recognize Their Environment and to Handle Objects
MANIPULATION

Development of Optical Communication Marks for Mobile Robots to Recognize Their Environment and to Handle Objects

Keiji Nagatani, Hiroyasu Sato, Hidenori Tasaka, Akio Gofuku, Yutaka TANAKA

发表年份
2005
引用次数
3

摘要

In an environment in which robots and human beings coexist, it is difficult for a mobile manipulator to grasp an object autonomously. Generally, the most difficult aspect of realizing such action is object recognition. Such an environment includes many of irregular shapes and sizes (because of unknown distance from sensors to each object) and blocking of vision sensors. Object’s abstraction and recognition thus remains far from practical use of current sensing technology. To solve the recognition problem, we developed marks recognized through optical communication rather than attempting to improve sensing technology. Marks using light-emitting diodes (LEDs) have the advantages of (1) being visible as far as they remain in sight, (2) having the properties of the target correctly using digital signals, and (3) localizing the mark position using stereovision. Once the marks are attached to several movable objects, a robot can find a location of a target object easily by communicating with the mark attached to the target object. Also, if the marks are located at known positions such as walls, a mobile robot can localize itself by detecting such marks. In this research, we developed an Intelligent Mark System (IMS) between a robot and marks which uses visible and infrared LEDs for communication. Using this system, we performed (1) object grasping tasks (the target is a small can with IMS) using a mobile manipulator, (2) localizing a mobile robot using IMS on several walls and (3) recognizing large objects (desks). In this paper, we explain an overview of communication between marks and robots, and we discuss task performance results for an autonomous mobile manipulator using IMS in a real environment. Also, we report the feasibility and limitations of our proposal.

关键词

GRASPRobotComputer visionMobile robotComputer scienceObject (grammar)Artificial intelligenceHuman–computer interaction

相关论文

查看 MANIPULATION 分类全部论文