Kinect-based human-following mobile robot

• Main Authors: Chao-Cheng Li, 李晁政
• Other Authors: Ching-Long Shih
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/8uryum

碩士 === 國立臺灣科技大學 === 電機工程系 === 105 === This thesis aims to an implement human-following robot equipped with Kinect sensor, so that the mobile robot can follow the specific person in the indoor environment. Kinect sensor was used to detect the skeleton of the human body. The operator was defined by the one who started with a default gestures, and the color of its upper outer garment was recorded. The color of the upper outer garment was defined as the following target, so that the robot would follow the one who dressed in the target color. And the relative distance and direction of angle to the target were calculated by the depth information and Kinect camera’s Cartesian coordinate. And an integrating LRF was used to avoid the obstacle from surroundings. In the experimental result, the mobile robot which controlled by closed loop PI controller could follow the target person automatically.