近年來視線追蹤技術的應用越來越廣，但是此技術大多應用於人臉偵測、人臉識別、人臉比對等用途上，在這些應用上皆需要極高的準確度，通常計算量會相當的大，能夠辨識的距離較近，所需要的影像解析度也較高。因此，如何使用簡單的計算簡化這些流程，並且使整個系統從原先的單人辨識提升至多人辨識，使其可以應用至更多的地方，便是一個很值得研究的議題。
本論文提出一個以頭部姿態角為基礎的多人視線判斷系統，第一步利用SeetaFace系統抓取臉部的81個關鍵點，第二步使用最小魏卡森迴歸分析這些關鍵點得到各個器官的特徵值，第三步利用這些特徵值，經由簡單的計算得到頭部姿態角，最後將頭部姿態角送入模糊規則庫系統，可獲得目標的視線方向。
本論文提出的方法主要是為了改善傳統視線偵測的方式，並且可以進行多人視線偵測，經由實驗統計證明，此方法有一定的準確度，可以對多個目標進行視線判斷。

In recent years, the application of gaze tracking technology has become more and more extensive, but this technology is mostly used in face detection, face recognition, face matching and other purposes. These applications require extremely high accuracy and usually require a considerable amount of calculation. The recognition distance is relatively short, and the required image resolution is also high. Under these conditions, the system usually targets a single person. Therefore, how to use simple calculations to simplify these processes and upgrade the entire system from the original single-person identification to multi-person identification, so that it can be applied to more places, is a topic worthy of research.
This thesis proposes a fuzzy rule-based gaze judgement system for multi-person gaze system based on facial features and head posture angles. The first step is to use the SeetaFace system to capture 81 key points of the face, and the second step is to use the Least Wilcoxon-based regression analysis to take the facial features. The key points are obtained by calculated, and the eigenvalue of each organ can be obtained through regression analysis. The third step is to use the eigenvalue from regression analysis to obtain the head posture angles through simple calculation, and finally send it to the fuzzy rule-based system for obtaining the sight direction of the targets.
The method proposed in this paper is mainly to improve the traditional way of sight detection, and can detect the sight of multiple people. Experiments have proved that this method has certain accuracy and can judge the sight of multiple targets.

第一章 緒論 1
第一節 前言 1
第二節 研究動機 1
第三節 論文回顧 3
第四節 研究方法 5
第五節 論文架構 7
第二章 相關研究方法 9
第一節 取得臉部關鍵點 9
第二節 迴歸分析 10
第三節 頭部姿態角判斷 12
第四節 模糊推論系統 13
第三章 視線偵測判斷系統 17
第一節 頭部關鍵點標註 17
第二節 迴歸分析 21
第三節 頭部姿態角判斷 35
第四節 模糊規則庫系統 49
第四章 實驗結果 57
第一節 實驗說明 57
第二節 實驗結果 57
第五章 結論與未來展望 79
第一節 結論 79
第二節 未來展望 79

D. A. Robinson, “A method of measuring eye movement using a scleral search coil in a magnetic field,” IEEE Transactions on Bio-medical Engineering, vol. BME-10, pp. 137-145, Oct. 1963.
A. Plotkin, O. Shafrir, E. Paperno, and D. M. Kaplan, “Magnetic eye tracking: a new approach employing a planar transmitter,” IEEE Transactions on Bio-medical Engineering, vol. 57, no. 5, pp. 1209-15, May 2010.
R. Barea, L. Boquete, and M. Mazo, “System for assisted mobility using eye movements based on electrooculography,” Neural Systems and Rehabilitation Engineering, pp.209-218, 2002.
K. Yamagishi, J. Hori, and M. Miyakawa, “Development of EOG-based communication system controlled by eight-directional eye movements,” in Proc. the 28th IEEE EMBS Annual International Conference, pp. 2574-2577, 2006.
A. Lopez, I. Rodriguez, F. J. Ferrero, M. Valledor, and J. C. Campo, “Low-cost system based on electro-oculography for communication of disabled people,” in Proc. Multi-Conference on Systems Signals & Devices (SSD) 2014 11th International, pp. 1-6, 2014.
L. Sun, Z. Liu, and M.-T. Sun, “Real time gaze estimation with a consumer depth camera,” Information Sciences, vol. 320, pp. 346-360, 2015.
P. M. Corcoran, F. Nanu, S. Petrescu and P. Bigioi, “Real-time eye gaze tracking for gaming design and consumer electronics systems,” IEEE Transactions on Consumer Electronics, vol. 58, no. 2, pp. 347-355, May 2012.
T. Sugita, S. Suzuki, J. Kolodko and H. Igarashi, “Development of Head-Mounted Eye Tracking System Achieving Environmental Recognition Ability,” in Proc. SICE Annual Conference 2007, pp. 1887-1891, 2007.
L.A. Zadeh, “Fuzzy Sets,” Information Control, vol. 8, pp338-353, 1965.
https://github.com/seetafaceengine/SeetaFace2(2020,July 17)
https://github.com/suyuan945/FaceLandmark1000(2020,July 17)
Stigler, Stephen M. “Gauss and the Invention of Least Squares.” The Annals of Statistics, vol. 9, no. 3, pp. 465–474,1981.
R.V. Hogg, J.W. McKean, and A.T. Craig, Introduction to Mathematical Statistics, 6th ed. Englewood Cliffs, NJ: Prentice-Hall, 2005.
蔡承翰，自構式最小魏卡森廣義的放射狀基函數類神經模糊推理系統之研究，碩士論文，國立東華大學電機工程學系，2008。https://hdl.handle.net/11296/2m8uxz。
鄭宇盛，基於最小魏卡森之前車距離偵測系統，碩士論文，國立東華大學電機工程學系，2018。https://hdl.handle.net/11296/h8jffv。
高子祺，應用於線上學習專注度評分之視線追蹤系統，碩士論文，國立東華大學電機工程學系，2018。https://hdl.handle.net/11296/tnjct8。
王宣竣，基於幾何人臉特徵之智慧型頭部姿態估測，碩士論文，國立交通大學電控工程研究所，2013。https://hdl.handle.net/11296/mw9pqr。
E. Murphy-Chutorian and M. M. Trivedi, “Head Pose Estimation in Computer Vision: A Survey,” in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 31, no. 4, pp. 607-626, April 2009, doi: 10.1109/TPAMI.2008.106.
孫宗瀛，楊英魁(2005)。Fuzzy 控制：理論、實作與應用(修訂版)。臺北市：全華。
N. Vadiee, R. Mohammad and T. Akbarzadeh, “Analogous fuzzy rule-based expert systems,” in Proc. IEEE 5th International Fuzzy Systems, 1996, pp. 1852-1857 vol.3, doi: 10.1109/FUZZY.1996.552679.
https://www.ndhu.edu.tw/var/file/0/1000/img/1627/20200529_2_4.jpg(2020/11/11)
https://www.ndhu.edu.tw/var/file/0/1000/img/1627/20200514_2_2.jpg(2020/11/11)
https://www.ndhu.edu.tw/var/file/0/1000/img/1627/20200514_2_3.jpg(2020/11/11)
https://www.ndhu.edu.tw/var/file/0/1000/img/1627/20201109_1_3.jpg(2020/11/11)
https://www.ndhu.edu.tw/var/file/0/1000/img/1627/20200615_1_2.jpg(2020/11/11)
https://www.ndhu.edu.tw/var/file/0/1000/img/1627/20201112_1_6.jpg(2020/11/11)
https://www.ndhu.edu.tw/var/file/0/1000/img/1627/20200825_2_3.jpg(2020/11/11)
https://www.ndhu.edu.tw/var/file/0/1000/img/1627/20200615_3_2.jpg(2020/11/11)
https://www.ndhu.edu.tw/var/file/0/1000/img/1627/20200903_1_5.jpg(2020/11/11)
https://www.ndhu.edu.tw/var/file/0/1000/img/1627/20201028_1_2.jpg(2020/11/11)
https://www.ndhu.edu.tw/var/file/0/1000/img/1627/20201023_2_2.jpg(2020/11/11)