• 沒有找到結果。

結論和未來展望

本論文製作一套具有三維方向頭動補償機制的眼動儀系統,使用一副頭戴式 眼動儀,包含一台實景攝影機和一台瞳孔攝影機,其中瞳孔攝影機用於拍攝瞳孔 影像,並使用自動二值化法和橢圓擬合取得瞳孔中心位置;實景攝影機拍攝使用 者前方影像,在螢幕上貼上四張圖卡,藉由辨識四張圖卡來預估螢幕三維位置。

在校正時,計算映射函數和以圖卡深度決定校正平面,使用時將瞳孔中心代入映 射函數求得在校正平面上的一點並計算視軸,依照圖卡位置計算螢幕平面,再計 算螢幕平面和視軸的交點作為凝視點,最後計算實景畫面中螢幕和實際螢幕間的 projective mapping 映射關係,將凝視點代入後顯示在螢幕上,達到頭動補償的 目的。

從實驗的結果和分析可以看出,本論文提出的頭動補償機制相較於先前的版 本(文獻[35]),在校正結果良好的情況下,頭部移動時可以更有效減少凝視點的 誤差,對於大幅頭動的補償也改善許多,但是從結果來看依然有進步空間,可以 從幾個角度來分析未來可以改良的方向:

(1) 映射函數

對於以映射函數來決定凝視點的眼動儀架構來說,映射函數是造成誤差的主 要來源之一,單從映射函數的角度來看,可以說當校正時計算出的映射函數精準 度越高時,本論文所使用的方法會更準確,但若是映射函數的精準度不佳,反而 會弄巧成拙導致誤差增加。因此若要以映射函數為基準預估凝視點,如何找出合 適的映射函數來進行凝視點預估是可以改進的方向。

(2) 硬體改良

在硬體部分也是可以改進的方向之一,例如本實驗使用的眼動儀是以鏡架的 方式固定在頭部,在使用時鏡架可能會晃動導致凝視點預估出現誤差,若能以更

解析度或是使用更小型的攝影機,對於凝視點的精準度和使用上的方便性都能有 所提升。

(3) 預估螢幕位置方式

本論文是以實景攝影機照射貼在螢幕上的圓形圖卡預估螢幕位置,但在預估 圖卡位置時必然會存在誤差,也會影響到預估出的凝視點誤差,因此如何更準確 的預估螢幕位置也是改良的方向之一。

(4) 眼睛位置的量測

本論文是以人工的方式大略量測使用者的眼睛位置,但以這種方式必定無法 準確量測到真正的眼睛位置,若能更準確量測到眼睛位置,或許可以更提升精準 度。

(5) 校正方式

以本論文的架構來說,雖然使用時可以頭部移動,但在校正時頭部是不能夠 移動的,若是校正時頭部移動則預估出的映射函數會出現誤差,雖然進行實驗時 頭部會靠在平台上固定,但是在一般使用時通常頭部會有些許的移動,因此如何 在頭部會些許移動的情況下進行準確的校正也是將來可以改進的方向之一。

參考文獻和資料

[1] http://www.open.ac.uk/about/campus/jennie-lee-research-labs-our-services/eye-tracker-s ervices

[2] Junghoon Park, Taeyoung Jung, Kangbin Yim, "Implementation of an EyeGaze Tracking System for the Disabled People.", 2015 IEEE 29th International Conference on

Advanced Information Networking and Applications, March 2015.

[3] Satoshi Handa, Yoshinobu Ebisawa, "Development of head-mounted display with eye-gaze detection function for the severely disabled.", 2008 IEEE Conference on Virtual

Environments, Human-Computer Interfaces and Measurement Systems, July 2008.

[4] http://www.vision-systems.com/articles/print/volume-18/issue-5/departments/snapshots/

disabled-users-navigate-daily-routines-with-eye-tracker.html

[5] Huey Edmund, The Psychology and Pedagogy of Reading(Reprint). MIT Press 1968(originally published 1908)

[6] Guy Thomas Buswell, How people look at pictures. A study of the psychology of

perception in art. The University of Chicago Press, 1935

[7] H. HARTRIDGE and L. C. THOMSON, "METHODS OF INVESTIGATING EYE

MOVEMENTS.", Br J Ophthalmol. 1948 Sep; 32(9): 581–591.

[8] Crane,H.D., Steele, C.M. (1985). "Generation-V dual-Purkinje-image eyetracker.",

Applied Optics 24 (4): 527–537.

[9] http://ppw.kuleuven.be/home/english/research/lep/resources/purkinje [10] http://www.biopac.com/curriculum/l10-electrooculogram-i/

[11] http://www.tobiipro.com/

[12] http://news.gpwb.gov.tw/mobile/news.aspx?ydn=026dTHGgTRNpmRFEgxcbfXeut3O MJo83bj7NmgTAG00AlkzL4T0uTu6lbNP6gA5vEt9TjkS3bTQAzJZMETETFsVmzZsh

Vk6QlgLvun%2BFJNo%3D

[13] Zhiwei Zhu and Qiang Ji, "Eye Gaze Tracking Under Natural Head Movements",2005

IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05), June 2005

[14] Chi-Wu Huang, Shih-Chen Tseng, Zong-Sian Jiang, Chun-Wei Hu, "Projective mapping compensation for the head movement during eye tracking", Consumer Electronics -

Taiwan (ICCE-TW), 2014 IEEE International Conference on, 2014 MAY

[15] Kai Han, Xuan Wang, Zili Zhang, Hainan Zhao, "A Novel Remote Eye Gaze Tracking Approach with Dynamic Calibration", Multimedia Signal Processing (MMSP), 2013

IEEE 15th International Workshop on, 2013

[16] D. H. Yoo and M. J. Chung, "A novel non-intrusive eye gaze estimation using cross-ratio under large head motion," Computer Vision and Image Understanding, vol. 98, pp. 25-51, Apr. 2005.

[17] F. L. Coutinho and C. H. Morimoto, "Augmenting the robustness of cross-ratio gaze tracking methods to head movement", in Proc. 2012 Symposium on Eye-Tracking

Research and Applications (ETRA'12), Santa Barbara, CA, USA, Mar. 2012, pp. 59-66.

[18] Zhiwei Zhu and Qiang Ji, "Novel Eye Gaze Tracking Techniques Under Natural Head Movement", IEEE Transactions on Biomedical Engineering (Volume:54,Issue: 12), 2007.

[19] E. D. Guestrin, M. Eizenman, "General Theory of Remote Gaze Estimation Using the Pupil Center and Corneal Reflections", IEEE Transactions on Biomedical Engineering, 2006

[20] https://www.qt.io/developers/

[21] http://opencv.org/

[22] Nobuyuki Otsu. "A threshold selection method from gray-level histograms.", IEEE Trans.

[23] Suzuki, S. and Abe, K., "Topological Structural Analysis of Digitized Binary Images by Border Following.", CVGIP 30 1, pp 32-46, 1985.

[24] http://user.engineering.uiowa.edu/~dip/LECTURE/Segmentation2.html [25] http://docs.opencv.org/trunk/de/dc7/fitellipse_8cpp-example.html#gsc.tab=0

[26] Andrew W. Fitzgibbon, Maurizio Pilu, Robert B. Fisher. "Direct Least Squares Fitting of Ellipses.", Pattern Recognition, Proceedings of the 13th International Conference . 1996 [27] https://en.wikipedia.org/wiki/HSL_and_HSV

[28] https://en.wikipedia.org/wiki/YCbCr

[29] Kenji Suzuki, Isao Horiba, Noboru Sugie,

"Linear-time connected-component labeling

based on sequential local operations.",

Computer Vision and Image Understanding, Volume 89, Issue 1, January 2003

[30] M. K. Hu, "Visual Pattern Recognition by Moment Invariants", IRE Trans. Info. Theory, vol. IT-8, pp.179–187, 1962

[31] J. Flusser, "On the Independence of Rotation Moment Invariants", Pattern Recognition, vol. 33, pp. 1405–1410, 2000

[32] J. Flusser and T. Suk, "Rotation Moment Invariants for Recognition of Symmetric Objects", IEEE Trans. Image Proc., vol. 15, pp. 3784–3790, 2006

[33] http://wenku.baidu.com/view/820e414d852458fb770b56db

[34] Paul S. Heckbert, "Projective mappings for image warping.", Image-Based Modeling and

Rendering(1999).

[38] Chi-Wu Huang, Wei-Chieh Tan, "An Approach of Head Movement Compensation when using a head mounted eye tracker", Consumer Electronics - Taiwan (ICCE-TW), IEEE, May 2016

自 傳

我的名字是唐唯傑,大學就讀於國立台灣師範大學電機工程學系,大學專題 主要是製作眼動儀的應用程式,包含眼動鍵盤、凝視軌跡紀錄和眼動繪圖及寫字 程式。研究所就讀於國立台灣師範大學電機工程學系碩士班,碩士班主要研究項 目為眼動儀系統的研究與開發和眼動儀頭動補償問題的研究,碩士論文為開發一 套具有三維方向頭動補償功能的眼動儀系統,針對使用眼動儀時頭部移動會造成 凝視點誤差的問題進行頭動補償修正誤差,改善實驗室之前眼動儀系統中的不足,

特別是針對瞳孔偵測和頭動補償機制做了許多改變和改進,實驗結果也顯示確實 有改善許多問題。

學 術 成 就

1. Chi-Wu Huang, Wei-Chieh Tan, "An Approach of Head Movement Compensation when using a head mounted eye tracker", Consumer Electronics - Taiwan (ICCE-TW), IEEE, May 2016