The design of intelligent interactive service robot

This study aims to develop the intelligent interactive multi-functional robot to provide companion and entertainment. To obtain the environmental information, we use Kinect depth image as the visual system platform and to perform the image processing operations. Then, the image processing results are applied to the robot behavior planning, integrated with omni-directional wheels, high power motors, FPGA and ARM-based development board as motion system. The lower body of the robot is constructed by four omni-directional wheels, and the upper body is in humanoid design. In this study, we use the information extracted from depth image stream of Kinect and integrate the robot localization system to build the real-time environment map and achieve the obstacle avoidance. In addition, we use Kinect skeleton detection and human facial expression recognition to accomplish the task of the intelligent interaction.