The proposed methods of this work is verified by experiment that they are indeed effective in reducing head motion error and have average error less than 1° and also
suitable for 2m usage. The first three methods rely on geometry information to compensate for head motion estimation deviation. Estimation of eye position is needed and might need additional device to point out the position information of eye in real time implementation. The last method, area coordinate mapping method need no additional information except for the initial gaze calibration process. Area coordinate mapping method might be weaker in theoretical model which involves many assumptions which make it inconvincible in theory whether it could handle all kind of estimation and head motion scenario robustly. Nevertheless, it is most appreciated for its easy setup and no additional hardware or calibration process needed. Geometry restoration method is the most complicated method since it needs more information such as camera and eye parameters which might need additional calibration. And this method could only use in nearly stationary head position since camera is not allowed to rotate in this method. The advanced geometry restoration method and distance normalization PCCR method is designed for a rotatable camera which is set to track the eye to the center of the captured image. These two methods are easier in whether calculation or calibration relative to
geometry restoration method and could allow camera rotation to keep eye in the image.
In general, these three methods are based on complete geometry model analysis and might be considered more reliable and robust to handle all kinds of system setting and head motion with a stable certain error. In theory, the advanced geometry restoration method could attain more accurate result close to stationary head position if the system and eye position is well calibrated and estimated.
However, the camera is currently manually set to collimate with cornea center, which is impossible for real time implementation. And more research data and experiment for more subjects are needed to give a further examination for these methods. In future, the camera is desirable to set on the pan tilt unit and tracking cornea center with corneal reflection position feedback. And also, the system speed could be further enhanced by inter frame pupil ROI tracking which uses the prior pupil position to assist the detection of later pupil.
REFERENCES
[1] S. P. Liversedge and J. M. Findlay, "Saccadic eye movements and cognition,"
Trends in Cognitive Science, vol. 4, no. 1, pp. 6–14, 2000.
[2] R. J. K. Jacob, "The use of eye movements in human computer interaction techniques: What you look at is what you get, " ACM Trans. Inf. Syst., vol. 9, no.
3, pp. 152–169, 1991.
[3] G. Underwood, "Cognitive Processes in Eye Guidance," Oxford Univ. Press, 2005.
[4] S. P. Liversedge and J. M. Findlay, "Saccadic eye movements and cognition,"
Trends in Cognitive Science, vol. 4, no. 1, pp. 6–14, 2000.
[5] M. F. Mason, B. M. Hood, and C. N. Macrae, "Look into my eyes: Gaze direction and person memory," Memory, vol. 12, pp. 637–643, 2004.
[6] R. J. K. Jacob, "The use of eye movements in human computer interaction techniques: What you look at is what you get," ACM Trans. Inf. Syst., vol. 9, no. 3, pp. 152–169, 1991.
[7] S. Zhai, C. H. Morimoto, and S. Ihde, "Manual and gaze input cascaded (magic) pointing," in Proc. ACM SIGHCI-Human Factors Comput. Syst. Conf., 1999, pp.
246–253.
[8] Z. Zhu and Q. Ji, "Eye and gaze tracking for interactive graphic display," Mach. Vis.
Application, vol. 15, no. 3, pp. 139–148, 2004.
[9] Rafael C. Gonzalez, Richard E. Woods, "Digital Image Processing", Prentice-Hall, Inc., Second Edition, 2002.
[10] J. Merchant, R. Morrissette, and J. Porterfield, "Remote Measurements of Eye Direction Allowing Subject Motion over One Cubic Foot of Space," IEEE Trans.
Biomedical Eng., vol. 21, no. 4, pp. 309-317, July 1974.
[11] N. Ramanauskas, "Calibration of video-oculographical eye-tracking system,"
Elektronika ir Elektrotechnika, vol. 72, pp. 65-68, 2015.
[12] Suwitchaya Rattarom, Nattapol Aunsri and Surapong Uttama, "Interpolation Based Polynomial Regression for Eye Gazing Estimation: A Comparative Study", 2015 IEEE International Conference on ECTI-CON.
[13] J. Coughlan, A. Yuille, C. English, and D. Snow, "Efficient Deformable Template Detection and Localization without User Initialization," Computer Vision and Image Understanding, vol. 78, no. 3, pp. 303-319, 2000.
[14] C. H. Morimoto and M. Mimica, "Eye gaze tracking techniques for interactive applications, " Comput. Vi. Image Understand, Special Issue on Eye Detection and Tracking, vol. 98, no. 1, pp. 4–24, 2005.
[15] Dong-Chan Cho and Whoi-Yul Kim, "Long-Range Gaze Tracking System for Large Movements", IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL.
[16] P. Viola and M. J. Jones, "Robust real-time face detection," International journal of computer vision, vol. 57, pp. 137-154, 2004.
[17] Zhiwei Zhu and Qiang Ji, "Novel Eye Gaze Tracking Techniques Under Natural Head Movement," IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 54, NO. 12, DECEMBER 2007.
[18] Dan Witzner Hansen and Qiang Ji, "In the Eye of the Beholder: A Survey of Models for Eyes and Gaze," IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL. 32, NO. 3, MARCH 2010.
[19] R.H.S. Carpenter, "Movements of the Eyes," Pion Limited, 1988.
[20] E.D. Guestrin and M. Eizenman, "General Theory of Remote Gaze Estimation Using the Pupil Center and Corneal Reflections," IEEE Trans. Biomedical Eng., vol. 53, no. 6, pp. 1124-1133, June 2006.
[21] Takehiko Ohno, Naoki Mukawa, Atsushi Yoshikawa "FreeGaze: A Gaze Tracking System for Everyday Gaze Interaction," Proceedings of the symposium on ETRA:
eye tracking research and applications symposium, pp.125-132, 2002.
[22] Anuradha Kar and Peter Corcoran, "A Review and Analysis of Eye-Gaze Estimation Systems, Algorithms and Performance Evaluation Methods in Consumer Platforms," Digital Object Identifier 10.1109/ACCESS.2017.2735633 [23] Alfred Yarbus, "Eye Movements and Vision," PLENUM PRESS NEW YORK 1967.
[24] S. Milekic, "The more you look the more you get: Intention-based interface using gaze-tracking," in Trant, J.(Des.) Museums and the Web 2002: Selected Papers from an Int. Conf., Archives and Museum Informatics, 2002, pp. 1–27.
[25] K. Hyoki, M. Shigeta, N. Tsuno, Y. Kawamuro, and T. Kinoshita, "Quantitative electro-oculography and electroencephalography as indexes of alertness,"
Electroencephalogr. Clinical Neurophysiol., vol.106, pp. 213–219, 1998.
[26] A. E. Kaufman, A. Bandopadhay, and B. D. Shaviv, "An eye tracking computer user interface," in Virtual Reality, 1993. Proceedings., IEEE 1993 Symposium on Research Frontiers in, pp. 120-121, 1993.
[27] D. A. Robinson, "A method of measuring eye movemnent using a scieral search coil in a magnetic field," IEEE Transactions on bio-medical electronics, vol. 10, pp.
137-145, 1963.
[28] A. Villanueva, R. Cabeza, and S. Porta, "Eye Tracking: Pupil Orientation Geometrical Modeling," Image and Vision Computing,vol. 24, no. 7, pp. 663-679, July 2006.
[29] A. Villanueva, R. Cabeza, and S. Porta, "Gaze Tracking System Model Based on Physical Parameter, " Int’l J. Pattern Recognition and Artificial Intelligence, 2007.
[30] T. Ohno, N. Mukawa, and A. Yoshikawa, "Freegaze: A Gaze Tracking System for
125-132, 2002.
[31] S.-W. Shih, Y.-T. Wu, and J. Liu, "A Calibration-Free Gaze Tracking Technique, "
Proc. 15th Int’l Conf. Pattern Recognition, pp. 201-204, 2000.
[32] Sheng-Wen Shih and Jin Liu, "A Novel Approach to 3-D Gaze Tracking Using Stereo Cameras," IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS—
PART B: CYBERNETICS, VOL. 34, NO. 1, FEBRUARY 2004.
[33] Craig Hennessey, Borna Noureddin and Peter Lawrence, "A Single Camera Eye-Gaze Tracking System with Free Head Motion," ETRA 2006, San Diego, California, 27–29 March 2006.
[34] GOSS, D. A., AND WEST, "Introduction to the optics of the eye. Butterworth-Heinemann," R. W. 2002.
[35] Chunfei Ma, Kang-A Choi, Byeong-Doo Choi and Sung-Jea Ko, "Robust remote gaze estimation method based on multiple geometric transforms," Optical Engineering 54(8), 083103 (August 2015).
[36] D. H. Yoo and M. J. Chung, "A novel non-intrusive eye gaze estimation using cross-ratio under large head motion,'' Comput. Vis. Image Understand., vol. 98, no. 1, pp.
25-51, Apr. 2005.
[37] I. F. Ince and J. W. Kim, "A 2D eye gaze estimation system with low resolution webcam images,'' EURASIP J. Adv. Signal Process., vol. 2011,no. 1, p. 40, Dec.
2011.
[38] W. Wang, Y. Huang, and R. Zhang, "Driver gaze tracker using deformable template matching,'' in Proc. IEEE Int. Conf. Veh. Electron. Safety, Jul. 2011, pp. 244-247.
[39] D. Beymer and M. Flickner, "Eye gaze tracking using an active stereo head," in Proc. Int. Conf. Computer Vision and Pattern Recognition, 2003, pp. 451–458.
[40] T. Ohno and N. Mukawa, "A free-head, simple calibration, gaze tracking system that enables gaze-based interaction," in Proc. Symp. Eye Tracking Res. Appl., 2004, pp. 115–122.
[41] C. A. Hennessey and P. D. Lawrence, "Improving the accuracy and reliability of remote system-calibration-free eye-gaze tracking," IEEE Trans.Biomed. Eng., vol.
56, no. 7, pp. 1891–1900, Jul. 2009.
[42] C.Hennessey and J. Fiset, "Long range eye tracking: Bringing eye tracking into the living room," in Proc. Symp. Eye Tracking Res. Appl., 2012, pp. 249–252.
[43] D. C. Cho,W. S.Yap, H. K. Lee, I. J. Lee, and W. Y.Kim, "Long range eye gaze tracking system for a large screen," IEEE Trans. Consum. Electron., vol. 58, no. 4, pp. 1119–1128, Nov. 2012.
APPENDIX
ACM Position 1 (0, 0, 1825), Movement (0, 0, 0).
Table A-1 ACM result for 16 testing points in position 1.
X=50 X=505 X=960 X=1415 X=1870
ACM Position 2 (0, 0, 1625), Movement (0, 0, -200).
Table A-2 ACM result for 16 testing points in position 2.
X=50 X=505 X=960 X=1415 X=1870
ACM Position 3 (0, 0, 2025), Movement (0, 0, 200).
Table A-3 ACM result for 16 testing points in position 3.
X=50 X=505 X=960 X=1415 X=1870
ACM Position 4 (0, -80, 1825), Movement (0, -80, 0).
Table A-4 ACM result for 16 testing points in position 4.
X=50 X=505 X=960 X=1415 X=1870
ACM Position 5 (0, -80, 1625), Movement (0, -80, -200).
Table A-5 ACM result for 16 testing points in position 5.
X=50 X=505 X=960 X=1415 X=1870
ACM Position 6 (0, -80, 2025), Movement (0, -80, 200).
Table A-6 ACM result for 16 testing points in position 6.
X=50 X=505 X=960 X=1415 X=1870
ACM Position 7 (-300, 0, 1825), Movement (-300, 0, 0).
Table A-7 ACM result for 16 testing points in position 7.
X=50 X=505 X=960 X=1415 X=1870
ACM Position 8 (-300, 0, 1625), Movement (-300, 0, -200).
Table A-8 ACM result for 16 testing points in position 8.
X=50 X=505 X=960 X=1415 X=1870
ACM Position 9 (-300, 0, 2025), Movement (-300, 0, 200).
Table A-9 ACM result for 16 testing points in position 9.
X=50 X=505 X=960 X=1415 X=1870
ACM Position 10 (-300, -80, 1825), Movement (-300, -80, 0).
Table A-10 ACM result for 16 testing points in position 10.
X=50 X=505 X=960 X=1415 X=1870
ACM Position 11 (-300, -80, 1625), Movement (-300, -80, -200).
Table A-11 ACM result for 16 testing points in position 11.
X=50 X=505 X=960 X=1415 X=1870
ACM Position 12 (-300, -80, 2025), Movement (-300, -80, 200).
Table A-12 ACM result for 16 testing points in position 12.
X=50 X=505 X=960 X=1415 X=1870
AGR Position 1 (0, 0, 1825), Movement (0, 0, 0).
Table A-13 AGR result for 16 testing points in position 1.
X=50 X=505 X=960 X=1415 X=1870
AGR Position 2 (0, 0, 1625), Movement (0, 0, -200).
Table A-14 AGR result for 16 testing points in position 2.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-15 Ideal “+” and estimated “o” gaze points w/o AGR in position 2.
AGR Position 3 (0, 0, 2025), Movement (0, 0, 200).
Table A-15 AGR result for 16 testing points in position 3.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-16 Ideal “+” and estimated “o” gaze points for AGR in position 3.
AGR Position 4 (0, -80, 1825), Movement (0, -80, 0).
Table A-16 AGR result for 16 testing points in position 4.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-19 Ideal “+” and estimated “o” gaze points w/o AGR in position 4.
AGR Position 5 (0, -80, 1625), Movement (0, -80, -200).
Table A-17 AGR result for 16 testing points in position 5.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-20 Ideal “+” and estimated “o” gaze points for AGR in position 5.
Fig. A-21 Ideal “+” and estimated “o” gaze points w/o AGR in position 5.
AGR Position 6 (0, -80, 2025), Movement (0, -80, 200).
Table A-18 AGR result for 16 testing points in position 6.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-23 Ideal “+” and estimated “o” gaze points w/o AGR in position 6.
AGR Position 7 (-300, 0, 1825), Movement (-300, 0, 0).
Table A-19 AGR result for 16 testing points in position 7.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-24 Ideal “+” and estimated “o” gaze points for AGR in position 7.
AGR Position 8 (-300, 0, 1625), Movement (-300, 0, -200).
Table A-20 AGR result for 16 testing points in position 8.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-27 Ideal “+” and estimated “o” gaze points w/o AGR in position 8.
AGR Position 9 (-300, 0, 2025), Movement (-300, 0, 200).
Table A-21 AGR result for 16 testing points in position 9.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-28 Ideal “+” and estimated “o” gaze points for AGR in position 9.
Fig. A-29 Ideal “+” and estimated “o” gaze points w/o AGR in position 9.
AGR Position 10 (-300, -80, 1825), Movement (-300, -80, 0).
Table A-22 AGR result for 16 testing points in position 10.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-31 Ideal “+” and estimated “o” gaze points w/o AGR in position 10.
AGR Position 11 (-300, -80, 1625), Movement (-300, -80, -200).
Table A-23 AGR result for 16 testing points in position 11.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-32 Ideal “+” and estimated “o” gaze points for AGR in position 11.
Fig. A-33 Ideal “+” and estimated “o” gaze points w/o AGR in position 11.
AGR Position 12 (-300, -80, 2025), Movement (-300, -80, 200).
Table A-24 AGR result for 16 testing points in position 12.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-35 Ideal “+” and estimated “o” gaze points w/o AGR in position 12.
DNP Position 1 (0, 0, 1825), Movement (0, 0, 0).
Table A-25 DNP result for 16 testing points in position 1.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-36 Ideal “+” and estimated “o” gaze points for DNP in position 1.
DNP Position 2 (0, 0, 1625), Movement (0, 0, -200).
Table A-26 DNP (4-34) result for 16 testing points in position 2.
X=50 X=505 X=960 X=1415 X=1870
Fig. A-37 Ideal “+” and estimated “o” gaze points for DNP (4-34) in position 2.