In everyday life environments like company offices, hospitals, classrooms, etc, tracking of movable objects such as chairs, note PCs is important to run smoothly the activity. We have developed an intelligent mobile robot that recognizes and counts the objects while moving in the environment. In addition, the robot places symbolic picture of the recognized objects in the pre-saved 2D map of the environment. The robot updates the database of objects in each room as objects are recognized. Therefore, it is possible to recognize the objects that moved from one room to another. The results show a good performance of the robot for object recognition and office inventory applications.

Kubo Shunsuke