Robust gaze-tracking method by using frontal-viewing and eye-tracking cameras

被引:16
作者
Cho, Chul Woo [1 ]
Lee, Ji Woo [1 ]
Lee, Eui Chul [2 ]
Park, Kang Ryoung [1 ]
机构
[1] Dongguk Univ, Dept Elect Engn, BERC, Seoul 100715, South Korea
[2] Sangmyung Univ, Dept Comp Sci, BERC, Seoul 110743, South Korea
关键词
gaze tracking; eye-tracking camera; frontal viewing camera; three rectangular regions; IRIS;
D O I
10.1117/1.3275453
中图分类号
O43 [光学];
学科分类号
070207 ; 0803 ;
摘要
Gaze-tracking technology is used to obtain the position of a user's viewpoint and a new gaze-tracking method is proposed based on a wearable goggle-type device, which includes an eye-tracking camera and a frontal viewing camera. The proposed method is novel in five ways compared to previous research. First, it can track the user's gazing position, allowing for the natural facial and eye movements by using frontal viewing and an eye-tracking camera. Second, an eye gaze position is calculated using a geometric transform, based on the mapping function among three rectangular regions. These are a rectangular region defined by the four pupil centers detected when a user gazes at the four corners of a monitor, a distorted monitor region observed by the frontal viewing camera, and an actual monitor region, respectively. Third, a facial gaze position is estimated based on the geometric center and the four internal angles of the monitor region detected by the frontal viewing camera. Fourth, a final gaze position is obtained by using the weighted summation of the eye and the facial gazing positions. Fifth, since a simple 2-D method is used to obtain the gazing position instead of a complicated 3-D method, the proposed method can be operated at real-time speeds. Experimental results show that the root mean square (rms) error of gaze estimation is less than 1 deg. (C) 2009 Society of Photo-Optical Instrumentation Engineers. [DOI:10.1117/1.3275453]
引用
收藏
页数:15
相关论文
共 28 条
  • [1] [Anonymous], 2006, Digital Image Processing
  • [2] Aotake Y., 1999, Proceedings of the IEEE International Conference on Systems, Man and Cybernetics, V3, P733, DOI DOI 10.1109/ICSMC.1999
  • [3] VISUALLY CONTROLLED GRAPHICS
    AZARBAYEJANI, A
    STARNER, T
    HOROWITZ, B
    PENTLAND, A
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1993, 15 (06) : 602 - 605
  • [4] Beach G, 1998, IEEE SYS MAN CYBERN, P4348
  • [5] CHO DH, 2005, P 4 INT C AS LANG PR, P61
  • [6] How iris recognition works
    Daugman, J
    [J]. IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, 2004, 14 (01) : 21 - 30
  • [7] Gemmell J., 2000, IEEE Multimedia, V7, P26, DOI 10.1109/93.895152
  • [8] HUMAN-COMPUTER INTERACTION USING EYE-GAZE INPUT
    HUTCHINSON, TE
    WHITE, KP
    MARTIN, WN
    REICHERT, KC
    FREY, LA
    [J]. IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS, 1989, 19 (06): : 1527 - 1534
  • [9] Ishii H, 2002, SICE 2002: PROCEEDINGS OF THE 41ST SICE ANNUAL CONFERENCE, VOLS 1-5, P1248, DOI 10.1109/SICE.2002.1195366
  • [10] Study on eyelid localization considering image focus for iris recognition
    Jang, Young Kyoon
    Kang, Byung Jun
    Park, Kang Ryoung
    [J]. PATTERN RECOGNITION LETTERS, 2008, 29 (11) : 1698 - 1704