隨行動裝置及無線網路的普及，智慧機器人的使用範疇從工廠擴展到商業應用甚至到個人應用，如何發展出適合人們使用的應用機器人成為熱門議題。其中，當使用者面向機器人操作時，使用者其頭部擺動資訊有相當價值，可用來附加更多功能。例如：視訊會議攝影機器人、主播播報攝影機器人及家用機器人等等。
因此本論文提出一套使用一般相機的頭部姿態即時追蹤系統。軟體部分，使用者面向相機，先運用Haar分類器技術偵測出人臉及鼻子特徵點，再利用光流法(Optical Flow)追蹤臉部特徵點，並使用疊代定比例正交投影演算法(Pose from Orthography and Scaling with Iterations, POSIT)求出特徵點三維位置，最後經過運算則可以獲得即時頭部擺向角度資訊(Real Time Head Pose Estimation)。硬體部分，搭載相機之機器人，在得到頭部擺向角度資訊後可使機器人即時追蹤使用者，讓使用者可隨心所欲轉動頭部，而機器人能移動至使用者臉部正前方。經實驗證明，本論文所提出的方法，於特定照明條件穩定下，在一幀時間內人臉轉動幅度20度以下可順利追蹤成功。

Following the popularization of mobile devices and wireless internet connection, the usage of intelligent robot has extended beyond industrial use into commercial, even personal applications, developing robots that suit general population’s needs is now a widely discussed topic. When a user is facing a robot while operating, the turning movement of head provides valuable information for further applications involving human-robot interactions, such as video conference robots, autonomous robot filming anchorman reporting news, or domestic robots, etc.
In this thesis we proposed a system for real-time head pose tracking system based on a single camera. Our software detects facial and nasal feature points of a human using Haar detector while the user is facing toward the onboard camera, tracks facial feature points using Optical Flow, later applies POSIT (Pose from Orthography and Scaling with Iterations, POSIT) algorithm to obtain 3-D position of feature points. Our hardware includes a robot with a camera onboard, after receiving turning information of the head movement the robot will begin tracking the user, therefore the user can move his head freely while the robot will stay directly in front of the user. Experimental results show that under stable lighting conditions the proposed method can successfully track a turning movement under 20 degree in a frame’s time.

[1] E. Murphy-Chutorian and M.M. Trivedi. Head pose estimation in computer vision: A survey. IEEE TPAMI. 2009
[2] T. F. Cootes, G. J. Edwards and C. J. Taylor. Active appearancemodels. IEEE TPAMI. 2001.
[3] T. Vatahska, M. Bennewitz and S. Behnke. Feature-based head pose estimation from images. In Humanoids. 2007.
[4] G. Fanelli, J. Gall, and L. Van Gool. Real time head pose estimation with random regression forests. In Computer Vision and Pattern Recognition (CVPR), 2011 IEEE Conference on, pages 617 –624, 2011.
[5] X. Yu, W. K. Chua, L. Dong, K. E. Hoe, and L. Li. Head pose estimation in thermal images for human and robot interaction. In Industrial Mechatronics and Automation (ICIMA), 2010 2nd International Conference on, vol. 2, pp. 698-701, 2010.
[6] Z. Kalal, K. Mikolajczyk, and J. Matas. Tracking-learning-detection. Pattern Analysis and Machine Intelligence, IEEE Transactions on,vol. 34, no. 7, pp. 1409–1422, 2012.
[7] K. Zhang, L. Zhang, and M.-H. Yang. Real-time compressive tracking. in Proceedings of the 12th European conference on Computer Vi-sion - Volume Part III, ECCV’12, (Berlin, Heidelberg), pp. 864–877,Springer-Verlag, 2012.
[8] G. Allen, R. Y. D. Xu, and J. S. Jin. Object Tracking Using CamShift Algorithm and Multiple Quantized Feature Spaces. In Proc. of Pan-Sydney Area Workshop on Visual Information Processing, VIP2003, 2003.
[9] H. C. Longuet-Higgins. Readings in computer vision: Issues, problems, principles, and paradigms. A Computer Algorithm for Reconstructing a Scene from Two Projections, pp. 61–62, San Francisco, CA, USA: Morgan Kaufmann Publishers Inc., 1987
[10] Z. Zhang. A flexible new technique for camera calibration. IEEE Transactions on Pattern Analysis and Machine Intelligence, Volume 22, Issue 11, pp. 1330-1334, 2000.
[11] D. F. DeMenthon and L. S. Davis. Model-based object pose in 25 lines of code. International Journal of Computer Vision, vol. 15, no. 1-2, pp. 123–141, 1995.
[12] P. Viola and M. Jones. Robust real-time face detection. In ICCV, volume 20(11), pages 1254-1259, 2001.
[13] M. Kearns. The computational complexity of machine learning. in Cambridge: MIT Press London, England: Massachusetts, 1990.
[14] Y. Freund and R. E. Schapire. A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting. Journal of Computer and System Sci-ences, vol. 55, pp. 119-139, 1997.
[15] B.D Lucas and T. Kanade. An iterative image registration technique with an application to stereo vision. Proceeding of the 7th International Joint Conference on Artificial Intelligence, pp.674–679, 1981.
[16] B.K.P. Horn and B. Schunk. Determining Optical Flow. Artificial Intelligence, vol. 17, pp. 185-203, 1981.
[17] W. Gander, G.H. Golub and R. Strebel. Least-Square Fitting of Circles and Ellipses. BIT, no. 43, pp. 558-578, 1994.
[18] P. L. Rosin. Ellipse fitting by accumulating five-point fits. Pattern Recognition Letters, 14:661–699,1993.
[19] 朱建亮。網路型機器人的設計與單相機LSD同步定位與建圖的實現。碩士論文。台北：國立臺灣科技大學機械工程研究所。2015。