Abstract
This paper describes a development of vision-based mobile robot, called Hyper Scooter. A user can ride on Hyper Scooter, and instruct it to act along his/her intention on-the-spot in a natural way.
Hyper Scooter is battery powered, carries a single driver, and has a system that integrates the correlation-based high speed tracking vision system and the user-interface to interact with the robot through visual information.
The features of our system are: (1) To make the mobile robot personal, we designed our system on the premise that a user rides on it. (2) the user requires no special knowledge. (3) Instruction and the execution are immediate supported by our high speed vision system.
The experimental result will show the feasibility of our system. The user can instruct Hyper Scooter to achieve autonomous complex trail replication instantly without explicit environmental model.
