1. 眼動儀與互動介面設計
廖文宏
6/26/2009

2. Outline Introduction Mobile eye tracker construction
Head-mounted eye tracker
Remote eye tracker
Head movement compensation
Human computer interface (HCI) applications
Conclusions

3. Introduction
An eye tracker is a device for measuring eye positions and eye movements.
The most popular variant uses video images from which the eye position is extracted.
Input source: visible spectrum vs. infrared

4. Eye Tracking at NCCU CS: Past
Scan path analysis using high speed iView X data.
Head mounted eye tracker for eye scrolling, eye gaming and eye typing.
scene camera
eye camera

5. Eye Tracking at NCCU CS: Present
Improve the pupil detection algorithm to alleviate corneal reflection problem.
Enhance the accuracy by compensating for head movement.
Construct and test a remote eye tracker.
More HCI applications using the remote eye tracker.
Use the eye tracking device to assist mobile user interface design.

6. System Architecture Eye image Calibration process Preprocessing
9 pairs
of points
Pupil detection
Calibration
Gaze point projection
Scene image

7. Pupil Detection Step 1: Feature detection Step 2: Noise removal
edge detection with constrains on neighboring dark points.
excludes edges created by bright spots
apply erosion
Step 2: Noise removal
Head mounted eye tracker
Remote eye tracker
Step 3: Fitting the ellipse using singular value decomposition

8. Allowing Head Movement: Head Mounted Eye Tracker
Use red markers on the LCD screen as the references for head movement calculation.
Required information:
Ratio between distance of makers in the projected image and in real: (dyWidthSclale,dyHeightScale)
Head movements: (difX,difY)
Correction made:

9. Compensate for head movement (error/standard deviation)
Experimental Results
Calibration point
Original error
Compensate for head movement (error/standard deviation) 1
98.17
(12.7) 2
137.54
55.89 (10.94) 3
108.20
29.69 (9.57) 4
76.30
28.79 (21.19) 5
113.30
22.49 (7.79) 6
117.86
34.83 (10.96) 7
116.53
33.31 (8.01) 8
112.56
16.03 (6.78) 9
146.73
26.49 (9.66)
1 cm = 38 Pixels

10. Allowing Head Movement: Remote Eye Tracker
Use makers placed on the glasses as references.
Requires an additional calibration step
Fixate on the same spot, and turn the head up, down, left and right.
Record the shift amount of the markers: (markXｉ,markYｉ) and movements of the pupil: (pupilXｉ,pupilYｉ)
Correction made:

11. Compensate for head movement (error/standard deviation)
Experimental Results
Calibration point
Original error
Compensate for head movement (error/standard deviation) 1
238.73
39.69 (24.6) 2
227.51
63.56 (22.82) 3
132.97
44.30 (21.49) 4
222.97
51.66 (30.94) 5
306.58
31.15 (16.71) 6
280.79
51.66 (21.83) 7
311.89
69.03 (41.57) 8
344.24
69.88 (31.71) 9
347.03
64.66 (24.41)

12. Demo: Head Mounted Eye Tracker

13. Demo: Web Browsing

14. Demo: Photo Viewing

15. Demo: Dynamic Scene

16. Demo: Remote Eye Tracker

17. Demo: Tic-Tac-Toe

18. Summary Enhance the reliability of pupil detection.
Improve the accuracy by compensating for head movements.
Promising results for both head mounted and remote eye trackers.
Interactive HCI applications.

19. Eye Tracking at NCCU CS: Future
Faster, more accurate eye tracking.
More intuitive calibration process.
Developing gaze-based digital interactive media.