- 學位論文
藉由實景影像特徵與模擬環境資訊之比對建立移動平台之行走路徑
Mobile Robot Self-Localization by Comparison of Real Image Feature and Environment Simulation
並列摘要
A novel self-localization is proposed for robots which using only an omnidirectional camera in this thesis. The method is first to generate a virtual space with a simulated omnidirectional camera to simulate robot equipped with an omnidirectional camera in the real world. Numerous vertical and horizontal lines are fictitious generated only in the virtual space and they are then projected to the simulated virtual omnidirectional image and often used the sphere camera model. We then combine the virtual omnidirectional image with the real one for matching the generated virtual lines with respect to the edges of the real omnidirectional image. Efficiently, matched lines are very precisely for estimating the pose or trajectory of the robot in the real world via the standard EPnP method. Experimental results validate the robot navigates in a trajectory and show the effectiveness of the proposed method in an indoor room.
參考文獻
延伸閱讀
- 鐘世凱（2004）。虛擬人物動畫行進動作模擬研究－即時最佳化步行路徑演算。藝術學報，(74)，113-123。https://doi.org/10.6793/JNTCA.200408.0113
- Wang, Y. S. (2016). 專為影像物件追蹤設計之背景相減與樣本比對模組平行化設計 [master's thesis, Feng Chia University]. Airiti Library. https://doi.org/10.6341/fcu.M0312375
- 王霆皓（2018）。以環境RGBD資訊進行輪腳複合機器人之地形判定導航與步態選擇〔碩士論文，國立臺灣大學〕。華藝線上圖書館。https://doi.org/10.6342/NTU201803465
- Yang, S. W. (2011). Mobile Robot Localization in Large-scale Dynamic Environments [doctoral dissertation, National Taiwan University]. Airiti Library. https://doi.org/10.6342/NTU.2011.02814
- Guo, D., Chen, H. W., & Liang, X. (2015). Adaptive Image-Based Leader-Follower Approach of Mobile Robot with Omnidirectional Camera. Journal of Applied Mathematics, 2015(), 113-124-010. https://doi.org/10.1155/2015/194251