Professional Documents
Culture Documents
ACKNOWLEDGEMENT
I would like to express my sincere gratitude and reverence to God Almighty, for guiding me through this seminar, making my endeavor an undiluted success. I am deeply indebted to our respected principal Dr.N.N.VIJAYA RAAGHAVAN for his timely advice and constant encouragement. I extend my sincere thanks to MRS.ANOOPA JOSE CHITTILAPPILLY, Head of Department of Applied Electronics & Instrumentation Engineering, for encouraging and aiding me throughout the seminar. I would like to express our heartfelt thanks to my seminar coordinator Ms.RESHMA RAMACHANDRAN, lecturer, AEI for her selfless support, understanding and involvement. I extend sincere and genuine appreciation to seminar guide Mrs.NEETHU SATHYAN, lecturer, AEI whose help throughout the seminar cannot be substituted by anything. In course of completion of the seminar I fortunate to receive the assistance of many faculties, friends and relatives who were extremely generous with their valuable suggestions, time and energy. I would like to thank all of them and recognize the fact that without them this seminar would have been inconceivable.
DEPT OF AEI
IESCE
ABSTRACT
The problem of eye gaze tracking has been researched and developed for a long time. Most of them use intrusive techniques to estimate the gaze of a person. This paper presents anon-intrusive approach for eye gaze tracker in real time with a simple camera. To track the eye gaze we have to deal with three principle problems: detecting the eye, tracking the eye and detecting the gaze of the eye on the screen where a user is looking at. In this paper we introduce the methods existed to solve these problems in the simple way and achieving high detection rate. The Eye-gaze System is a direct-select vision-controlled communication and control system. It was developed in Fairfax, Virginia, by LC Technologies, Inc., This system is mainly developed for those who lack the use of their hands or voice. Only requirements to operate the Eye-gaze are control of at least one eye with good vision & ability to keep head fairly still. Eye-gaze Systems are in use around the world. Its users are adults and children with cerebral palsy, spinal cord injuries, brain injuries, ALS, multiple sclerosis, brainstem strokes, muscular dystrophy, and Werdnig-Hoffman syndrome. Eye-gaze Systems are being used in homes, offices, schools, hospitals, and long-term care facilities. By looking at control keys displayed on a screen, person can synthesize speech, control his environment (lights, appliances, etc.), type, operate a telephone, run computer software, operate a computer mouse, and access the Internet and e-mail. Eye-gaze Systems are being used to write books, attend school and enhance the quality of life of people with disabilities all over the world.
DEPT OF AEI
IESCE
TABLE OF CONTENTS
CHAPTER
ACKNOWLEDGMENT ABSTRACT TABLE OF CONTENTS LIST OF FIGURES LIST OF TABLES
PAGE No.
i ii iii iv v
1. INTRODUCTION 2. WORKING 2.1 WORKING OF EYE-GAZE SYSTEM 2.2 TRACKING OF EYE MOVEMENT 2.3 LONGEST LINE SCANNING 2.4 OCEM 2.5 ESTIMATION OF GAZING 2.6 EYE DETECTING 2.7 ESTIMATION ALGORITHMS 3. OPERATIONAL REQUIREMENTS 3.1 LOW AMBIENT INFRARED LIGHT 3.2 EYE VISIBILITY 3.3 GLASSES & CONTACT LENSES 4. EYE GAZE PERFORMANCE SPECIFICATIONS 5. MENU OF EYE GAZE SYSTEM
01 02
10
12 14 18 19 20 25 31 26
6. SKILLS NEEDED BY USERS 7. EYE-GAZE TRACKER OUTPUT 8. DEVELOPMENTS IN EYE-GAZE SYSTEM 9. ADVANTAGES OF EYE GAZE SYSTEM 10. PERSONAL CONTRIBUTIONS AND VIEWS 11. CONCLUSION
DEPT OF AEI
IESCE
REFERENCES
22
LIST OF FIGURES
No. 1. 2. 3.
4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
NAME
Combined Eye-Gaze System Working Diagram of an Eye-Gaze System Detecting of Eye Pupil Principles of LLS Matching Process in OCEM Reference Model: 2D Simple Mark Two examples of positive image Geometry around Eye-Gaze Practical Eye Gaze System Main Menu Telephone Menu Typewriter Menu Run PC Menu Light & Appliances Menu Eye-Gaze System in Wheel Chair Computer Interaction Eye-Gage Video Streaming
PAGE No.
02 03 04 05 06 06 07 08 11 14 14 15 16 17 20 21 24
DEPT OF AEI
IESCE
LIST OF TABLES
No.
1
NAME
Accuracy
PAGE No.
12
DEPT OF AEI
IESCE
1. INTRODUCTION
Most of us blessed to operate the computer with ease using our hands. But there are some who cant use their hands and for them the voice guided systems have been in use for quite some time now. But what about paralytic patients with no mobility and speech? Even when their brains are functional and they are visually and aurally blessed to know what is going around. So shouldnt they be able to effectively use their intelligence and stay employed? Now they can with the Eyegaze communication system. Detecting of eye gaze is used in a lot of human computer interaction applications. Most of them use intrusive techniques to estimate the gaze of a person. For example, user has to wear a headgear camera to fix the position of their eyes with the view of screen on the camera, or use an infrared light on camera to detect the eye. In this paper, I introduce a nonintrusive approach which is very cheap solution to detect the eye gaze with a camera simple, user does not have to wear the headgear or using any expensive equipment. The Eye-gaze System is a direct-select vision-controlled communication and control system. It was developed in Fairfax, Virginia, by LC Technologies, Inc., This system is mainly developed for those who lack the use of their hands or voice. Only requirements to operate the Eye-gaze are control of at least one eye with good vision & ability to keep head fairly still. Eye-gaze Systems are in use around the world. Its users are adults and children with cerebral palsy, spinal cord injuries, brain injuries, ALS, multiple sclerosis, brainstem strokes, muscular dystrophy, and Werdnig-Hoffman syndrome. Eye-gaze Systems are being used in homes, offices, schools, hospitals, and long-term care facilities. By looking at control keys displayed on a screen, person can synthesize speech, control his environment (lights, appliances, etc.), type, operate a telephone, run computer software, operate a computer mouse, and access the Internet and e-mail. Eye-gaze Systems are being used to write books, attend school and enhance the quality of life of people with disabilities all over the world.
DEPT OF AEI
IESCE
2. WORKING
2.1 Working of Eye Gaze System
As a user sits in front of the Eye-gaze monitor, a specialized video camera mounted below the monitor observes one of the user's eyes. Sophisticated imageprocessing software in the Eye-gaze System's computer continually analyzes the video image of the eye and determines where the user is looking on the screen. Nothing is attached to the user's head or body.
Fig.2.1 Combined Eye-Gaze System In detail the procedure can be described as follows: The Eye-gaze System uses the pupil-center/corneal-reflection method to determine where the user is looking on the screen. An infrared-sensitive video camera, mounted beneath the System's monitor, takes 60 pictures per second of the user's eye. A low power, infrared light emitting diode (LED), mounted in the center of the camera's lens illuminates the eye. The LED reflects a small bit of light off the surface of the eyes cornea. The light also shines through the pupil and reflects off of the retina, the back surface of the eye, and causes the pupil to appear white. The bright-pupil effect enhances the camera's image of the pupil and makes it easier for the image processing functions to locate the center of the pupil.
DEPT OF AEI
IESCE
The computer calculates the person's gaze point, i.e., the coordinates of where he is looking on the screen, based on the relative positions of the pupil center and corneal reflection within the video image of the eye. Typically the Eye-gaze System predicts the gaze point with an average accuracy of a quarter inch or better.
Fig.2.2 Working Diagram of an Eye-Gaze System Prior to operating the eye tracking applications, the Eye-gaze System must learn several physiological properties of a user's eye in order to be able to project his gaze point accurately. The system learns these properties by performing a calibration procedure. The user calibrates the system by fixing his gaze on a small yellow circle displayed on the screen, and following it as it moves around the screen. The calibration procedure usually takes about 15 seconds, and the user does not need to recalibrate if he moves away from the Eye-gaze System and returns later.
can hardly be differentiated from the iris in the captured images. If the image is captured from close range, then it can be used to detect the pupil even under ordinary lighting conditions.
Fig.2.3 Detecting of Eye Pupil It was decided to track the iris for this reason. Due to the fact that the sclera is light and the iris is dark, this boundary can easily be optically detected and tracked. It can be quite appropriate for people with darker iris color (for instance, Asians). Young has addressed the iris tracking problem using a head-mounted camera.
DEPT OF AEI
IESCE
Fig 2.4: Principles of LLS Searching and decision after edge detection enhances computational efficiency. Except when preprocessing fails, it computes the center of the iris quite accurately. But it is sensitive to distribution of edge pixels.
DEPT OF AEI
IESCE
10
Fig2.5: Reference Model: 2D Simple Mark It cannot offer any orientation information at all, because it is like a small spot. Nevertheless, it can compensate for slight head movement.
DEPT OF AEI
IESCE
11
Fig 2.6: Two examples of positive image We get images from a lot of sources and select the eye manually. Then we use the create samples utility in Open CV to create training samples and using haartraining utility to train the classifier. It takes about five days to train our classifier with machine Pentium(R) 4 CPU 3.00GHz. If we use this classifier to tracking the eye in the real time, it will be very heavy, because it has to search the eye in all frames of camera. So we use this classifier to detect the eye in the first frame and then use another method to track the eye for all the next frames.
12
geometry of the eye gaze discussed in the previous section. Adaptive Estimation determines the eye-gaze with the help of displacements in the image. Regardless of which of these techniques is actually employed, image-to-screen mapping requires that the system be initialized first. It should be calibrated while in use. During initialization, the subject intentionally gazes at predefined screen points. From the resulting eye movement data, other gazing points can be estimated. During the calibration, because subject moves continuously, changes in the parameters (such as the radius or iris, the distance, or the head position arising due to subject movements) are incorporated in the estimation process, thereby reconstructing the parameter set.
S=k {
Fig 4: Geometry around Eye-Gaze During initialization, the value of k is expected to be different depending on the direction towards each predefined screen points. The different value of k can be
DEPT OF AEI IESCE
13
computed at this initialization stage. The value S refers to the gazing point. The situation is the same as in the initialization step:
DEPT OF AEI
IESCE
14
3. OPERATIONAL REQUIREMENTS
3.1 Low Ambient Infrared Light
There must be low levels of ambient infrared light falling on the subject's eye. Stray IR sources obscure the lighting from the Edge Analysis System's light emitting diode and degrade the image of the eye. The sun and incandescent lamps contain high levels of infrared light. The environment may be brightly illuminated with lights such as fluorescent or mercury-vapor which do not emit in the infrared region of the spectrum. The Edge Analysis System also works well in the dark.
DEPT OF AEI
IESCE
15
Fig 3.1: Practical Eye Gaze System Soft contact lenses that cover all or most of the cornea generally work well with the Edge Analysis System. The corneal reflection is obtained from the contact lens surface rather than the cornea itself. Small, hard contacts can cause problems, however, if the lenses move around considerably on the cornea, and the corneal reflection moves across the discontinuity between the contact lens and the cornea.
DEPT OF AEI
IESCE
16
0.25 inch
0.18 degree
0.06 inch
asymmetries of the pupil opening about the eye's optic axis, and astigmatism. They are constant from frame to frame and cannot be reduced by averaging or smoothing. Frame-to-frame variations result from image brightness noise and pixel
position quantization in the camera image and may be reduced by averaging or smoothing.
4.2 Speed
Sampling rate: 60 MHz
17
Gaze Cone diameter: 80 Degrees The lower 15 degrees of the gaze cone, however, are generally clipped due to the upper eyelid blocking the corneal reflection when the eye is looking down below the camera.
In fixed-camera Edge Analysis Systems, the eye must remain within the field of view of the camera. However, if the subject moves away from the camera's field of view, eye tracking will resume once he returns to a position where his eye is again visible to the camera.
DEPT OF AEI
IESCE
18
19
20
For children, two new Eye-gaze programs have been added to the Eye-gaze System. Both run with the Second PC option. Eye Switch is a big, basic on-screen switch to run "cause & effect" software programs on a Second PC. Simple Mouse is an easy mouse control program to provide simplified access to educational software on a Second PC.
5.5 Television
Television programs can be displayed directly on the desktop Eye-gaze System screen. On-screen volume and channel controls provide independent operation (Not available on the Portable Eye-gaze System.).A web browsing system using eye-gaze input. Recently, the eye-gaze input system was reported as a novel human-machine interface. We have reported a new eye gaze input system already. It utilizes a personnel computer and a home video camera to detect eye-gaze under natural light. In this paper, we propose a new web browsing system for our conventional eye-gaze input system.
21
DEPT OF AEI
IESCE
22
DEPT OF AEI
IESCE
23
Eye-Found Flag:
Gaze point:
Pupil Diameter:
Synchronization Counter:
DEPT OF AEI
IESCE
24
25
appreciable. Rather than applying a moving average smoothing filter, we dealt with this issue by finding a good compromise between the tolerance and the index reference parameters. Larger values of the tolerance parameter reduce the flickering effects, but also reduce the resolution of the Visual Mouse. Smaller values of the index reference parameter will generate the mouse click more quickly, which will decrease the effect of flickering. We found our trade-off values of these two system parameters to be robust for the applications to be described below. However we do not exclude the possibility that filtering of the incoming eye-gaze coordinate data stream could well lead to a more automated approach. We are currently studying the statistical properties of empirical eye-gaze coordinate data streams with such issues in mind.
Fig 8.2:Computer Interaction A direct implication of reducing the resolution of the Visual Mouse is as follows. The Visual Mouse works very well when hot links with big icons are involved. Dealing with smaller clickable icon links, however, is troublesome. An example of the latter was our attempt to use the Back button on a regular web browser window, in the context of web surfing. The Back button proved to be too small for the Visual Mouse to operate effectively. A second issue addressed was related to the accuracy of the calibration procedure. The Procedure for calibrating the PC screen for each subject and session used nine points in order to calculate the mapping that relates the subjects angle of gaze with positional coordinates on the approximately planar PC monitor screen. It is crucial to achieve good calibration of these nine points for positional accuracy of subsequent subject eye-gaze location. Notwithstanding the accuracy with which this is
DEPT OF AEI IESCE
26
done, there is some decrease of accuracy whenever the subject looks at the PC screen at locations away from the calibration points. One way to avoid the decrease in accuracy in defining eye-gaze positions on a planar screen is to use a greater number of calibration points. A seventeen-point calibration is possible, and will give better results, but it requires appreciably more time to carry out. In summary, we can gain in accuracy of pinpointing eye-gaze locations at the expense of time and effort (and hence subject fatigue) taken in calibration. All results reported on below used nine point calibration which provided an adequate trade-off between the subjects work on calibration, and the positional precision of the data consequently obtained.
27
A medical example is accessible at http://strule.cs.qub.ac.uk/imed.html. This image has been compressed from 8.4 MB to 568 kB, and again decompression (and format conversion) of limited area blocks is carried out effectively in real time. The compression method used in this case is rigorously loss-less, and supports grayscale images. The multi-resolution transform used is a pyramidal median transform. Further details on the compression algorithms used can be found in Starck et al. (1996, 1998), Louys et al. (1999a, 1999b), Murtagh et al. (1998, 2001a, 2001b). The eye-gaze control system can be used to operate these web-based demonstrations. The block sizes are sufficiently large that no precision problems are encountered with these types of applications. General context for such scientific and medical applications is as follows. New display and interaction environments for large scientific and medical images are needed. With pixel dimensions up to 16,000 x 16,000 in astronomy, which is the case of detectors at the CFHT (Canada-France-Hawaii Telescope, Hawaii) and the UKs Vista telescope to be built at the European Southern Observatorys facility at Cerro Paranal, Chile, it is clear that viewing navigation support is needed. Even a digitized mammogram in telemedicine, of typical pixel dimensions 4500 x 4500, requires a display environment. Our work concerns therefore both image compression, and also a range of other allied topics progressive transmission, views based on resolution scale, and quick access to full-resolution regions of interest. Ongoing work by us now includes the following goals: (i) better design of web-based display, through enhancing these demonstrators (including more comprehensive navigation support for large image viewing, and a prototyping of eyegaze controlled movement around three-dimensional scenes); and (ii) support for further Visual Mouse interaction modes. Chief among interaction modes is a back or return action based on lack of user interest, expressed as lack of gaze concentration in a small region.
28
sufficiently long on one of the panels, the video presentation is displayed in the panel. If the observers interest remains as measured by eye-gaze on the panel, the video continues to display. At any time, the observers interest may wander. If his or her eye dwells sufficiently long on another panel then the previously playing video is replaced with a name-plate, and a video stream now plays in the new panel. An HTML OBJECT element is used to insert an ActiveX component into the HTML document as well as all of the necessary information to implement and run the object.
DEPT OF AEI
IESCE
29
DEPT OF AEI
IESCE
30
DEPT OF AEI
IESCE
31
11. CONCLUSION
Today, the human eye-gaze can be recorded by relatively unremarkable techniques. This thesis argues that it is possible to use the eye-gaze of a computer user in the interface to aid the control of the application. Care must be taken, though, that eye-gaze tracking data is used in a sensible way, since the nature of human eyemovements is a combination of several voluntary and involuntary cognitive processes. The main reason for eye-gaze based user interfaces being attractive is that the direction of the eye-gaze can express the interests of the user-it is a potential porthole into the current cognitive processes-and communication through the direction of the eyes is faster than any other mode of human communication. It is argued that eyegaze tracking data is best used in multimodal interfaces where the user interacts with the data instead of the interface, in so-called non-command user interfaces.
DEPT OF AEI
IESCE
32
REFERENCES
(1)LAURENCE R. YOUNG AND DAVID SHEENA, Survey of Eye Movement Recording Methods, Behavior Research Methods and Instrumentation, Vol. 7, No. 5, pp. 397-429, 1975. (2)ARIE E. KAUFMAN, AMIT BANDOPADHAY, BERNARD D. SHAVIV, An Eye Tracking Computer User Interface , Research Frontier in Virtual Reality Workshop Proceedings, IEEE Computer Society Press, pp. 78-84. October 1993, (3)GLENN A. MYERS, KEITH R. SHERMAN, AND LAWRENCE STARK, Eye Mornitor, IEEE Computer Magazine, Vol. March, pp. 14-21, 1991. YOSHINOBU EBISAWA, Improved Video-Based Eye- Gaze Detection Method, IEEE IMTC 94, Hamamatsu, May, 1998. (4)THOMAS E. HUTCHINSON, K. PRESTON WHITE, JR., WORTHY N. MARTIN, KELLY C. REICHERT, AND LISA A. FREY, Human-Computer Interaction Using Eye- Gaze Input, IEEE Trans. on Systems, Man, and Cybernetics, Vol. 19, No. 6, pp. 1527-1534, 1998. (5)C. COLOMBO, S. ANDRONICO, AND P. DARIO, Prototype of a Vision-Based Gaze-Driven Man-Machine Interface, Proceedings IEEE/RSJ International Conference on Intelligent Robots and Systems, August, 1995. (6)CHRISTOPHE COLLET, ALAIN FINKEL, AND RACHID GHERBI, CapRe: A Gaze Tracking System in Man- Machine Interaction, Proceedings of IEEE International Conference on Intelligent Engineering Systems, September, 1997. (7)BAOSHEN Hu AND MINGHUA QIU, A New Method for Human-Computer Interaction by using Eye Gaze, Proceedings of IEEE International Conference on Systems, Man and Cybernetics, October, 1994. (8)RAINER STIEFELHAGEN, Gaze Tracking for Multimodal Human-Computer Interaction, Diplomarbeit, Universitiit Karlsruhe, September, 1996. (9)SHUMEET BALUJA AND DEAN POMERLEAU, Non- Intrusive Gaze Tracking Using Artificial Neural Networks, CMU Technical Report, CMU-CS-94-102, School of Computer Science, Carnegie Mellon University,January, 1994. (10)PHILIPPE BALLARD AND GEORGE C. STOCKMAN, Computer Operation via Face Orientation, Pattern Recognition, Vol.l. Conference A: Computer Vision and Applications, Proceedings., 11th IAPR International Conference, 1992.
DEPT OF AEI
IESCE
33
(11)A. H. GEE AND R. CIPOLLA, Determining the Gaze of Faces in Images, Technical Report, CUED/FINFENG/ TR 174, Department of Engineering, University of Cambridge, March, 1994. (12) P. Viola, M. Jones, Rapid Object Detection using a Boosted Cascade of Simple Features , IEEE vol. 2, 2001. (12) R. Lienhart, J. Maydt, An Extended Set of Haar-like Features for Rapid Object Detection, vol. 1, no. 1, pp. 900-903, 2002. (13) J. Bouguet, Pyramidal Implementation of the Lucas Kanade Feature Tracker Description of the algorithm, Intel Corporation, Microprocessor Research Labs, OpenCV Document, 1999 (14) C. Rasmussen, C. Williams, and I. Books24x7, Gaussian Processes for Machine Learning. Springer, 2006.
DEPT OF AEI
IESCE
34