碩士 / 國立臺北科技大學 / 電機工程系所 / 102 / In this thesis, a real-time map conversion strategy is proposed to generate a 3D navigation map based on the information explored by a scout robot in unknown environment. The 3D navigation map is further converted to a 2D navigation map based on which navigation trajectories for a blind mobile robot without onboard camera can be determined. According to the desired trajectory, a visual servo controller can be designed to drive the blind robot to the destination in real-time. The proposed system is composed of a set of surveillance cameras mounted on the corners of the ceiling capable of observing the blind robot with tilted view. Three visual servo controllers are proposed to drive the mobile robot to follow the desired trajectory by placing an appropriate surveillance camera in the feedback loop. Meanwhile, the estimation of the robot pose is improved by two monocular vision-based correction methods. When the blind robot is in the overlapped fields of view of two surveillance cameras, the system can switch to the next camera and continue executing the navigation task based on the transformation between two cameras. The proposed algorithms can be employed to construct navigation maps and drive the mobile robot to the desired position efficiently. Successful simulation and experimental results have validated the feasibility and effectiveness of the proposed methods.
| Identifer | oai:union.ndltd.org:TW/102TIT05442111 |
| Date | January 2014 |
| Creators | Huan-Chen Ling, 凌歡辰 |
| Contributors | Wen-Chung Chang, 張文中 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 72 |
Page generated in 0.0083 seconds