Visual feedback control of a mobile robot for mechatronics education

Fusaomi Nagata, Toshiyuki Tatai, Mamadou Ngom, Akimasa Otsuka, Maki K. Habib, Keigo Watanabe

Research output: Chapter in Book/Report/Conference proceedingChapter


Recently, visual feedback control system is becoming more attractive for mechatronics education due to the development of RGB-D cameras such as Kinect and Xtion. In this paper, an example of a simple visual feedback control system of a mobile robot with an axis-symmetric shape is introduced for mechatronics education which has to be demonstrated within a time limit of a lecture. Positions of a robot in image plane and projected plane can be calculated by referring to RGB stream and depth stream obtained from Xtion camera, respectively. As the first simple exercise, a virtual barrier fence is designed, so that a mobile robot can move within the virtual fence even without a real one. In addition, if a mobile robot has an axis-symmetric shape, e.g., circle, from the top view, it is difficult for a vision system to identify the orientation of the robot in the coordinate system. Another exercise is introduced to deal with an orientation following control using a forward direction vector. The forward direction vector can be calculated frompoint cloud data obtained by makingthe robot move forward for a short distance, e.g., 30 mm, every dynamic sampling period. The effectiveness and usability of the presented work is demonstrated experimentally.

Original languageEnglish
Title of host publicationComputer Vision and Simulation
Subtitle of host publicationMethods, Applications and Technology
PublisherNova Science Publishers, Inc.
Number of pages13
ISBN (Electronic)9781634858038
ISBN (Print)9781634857901
Publication statusPublished - Jan 1 2016

ASJC Scopus subject areas

  • Economics, Econometrics and Finance(all)
  • Business, Management and Accounting(all)


Dive into the research topics of 'Visual feedback control of a mobile robot for mechatronics education'. Together they form a unique fingerprint.

Cite this