RESNA Gaze Tracking System for Enhanced Human-Computer Interaction

Size: px

Start display at page:

Download "RESNA Gaze Tracking System for Enhanced Human-Computer Interaction"

Ashlynn Young
5 years ago
Views:

1 RESNA Gaze Tracking System for Enhanced Human-Computer Interaction Journal: Manuscript ID: Submission Type: Topic Area: RESNA 2008 Annual Conference RESNA-SDC Student Design Competition Computer Applications & Communication (CAC)

2 Page 1 of 7 RESNA Paper Type: U, Topic Area: CAC Gaze Tracking System for Enhanced Human-Computer Interaction Michael J. Lenisa, Breanna Heidenburg, Daniel Wentzel, Dr. Aleksander Malinowski ABSTRACT Eye based interface systems for computers have existed for several years, but have not come into widespread use due to the high cost of the associated technology and devices. Recent developments in the area of high grade consumer cameras with small form factors have created devices which allow for low cost gaze tracking research to be conducted. Previous design work done in this field has often neglected to provide feedback to the user, thereby limiting the potential uses of those systems. Results have improved upon previous best tracking accuracies, as well as achieving a speed increase over conventional mouse based systems. A Head Mounted Display has been integrated into the system, allowing full hands free use and mobility. Continuing advances in this field, such as those shown in this research, will rapidly lead to the designed system being viable for real world assistive technology use. KEYWORDS human interface device; hands free operation; gaze tracking; eye tracking BACKGROUND Gaze tracking has been used for many decades as a tool to study human cognitive process including reading, driving, watching commercials, and other activities (1). With the advent of personal computers, potential integration of such systems has been considered only recently in 1991 by Jacob (2). Successful attempts to employ gaze tracking as user interface were made to allow users with movement disabilities to type by looking at virtual keyboard (3), or for mouse pointer control (4). Systems of this kind have also been used by individuals without any disabilities to enhance performance (5, 6). The continuing absence of consumer-grade eye-tracking human computer interface technology is result of the high price of eye tracking technology and intrusiveness of such systems, despite the fact that technologies allowing so have existed for many years (7). Only recently relatively low cost systems have been investigated, for example by Li et al in 2006 (8-10). A full technical description of this project, along with complete source code is available at the project website (11). Also, this research has been slated for publishing at the HIS conference in May The design work described herein shows the accuracy that can be obtained using modern, low-cost systems. The two most significant contributions of this research are the use of blob detection in the visible spectrum instead of circular shape in the infrared spectrum, and use of higher dimensional polynomials to calibrate and map the detected gaze position to the position on the computer screen. DESIGN AND DEVELOPMENT The design of this gaze tracking system called for an image capture device working in the visible spectrum to capture images of the user s eye. In order to provide reasonable resolution, the camera is placed slightly below the user s eye, positioned facing up towards the eye. In the design process it was found that the physiology of the human eye does not allow for easy shape detection of the pupil, as every user s eye is different, as well as slight shape changes of the pupil at various deflection angles. Therefore, image processing techniques such as contrast stretching,

3 RESNA Page 2 of 7 binary thresholding, and blob detection are used to correctly identify the pupil. After first identifying the center of a user s pupil, the Cartesian coordinates of the center are passed through a mapping algorithm to determine a cursor position on a head mounted display. After the user has properly calibrated the system, a set of constants is derived which are used in a point transformation algorithm. The point transformation algorithm accepts the set of calibration constants, as well as the location of the center of the pupil within the image. The output of this function is the coordinate position of a cursor on the screen representing where the user s gaze is focused. This process allows for gaze-based cursor positioning control, as well as action based interaction, such as clicking. Clicking is achieved via dwell time, wherein past positions of the user s gaze are used to determine if the user has been focusing in a small region for a given period of time. The size of the region and the dwell time required are variable parameters within the system. The system accuracy is limited by inherent properties of the human eye. Examples of problems contributing to inaccuracy include: saccadic eye movements, ocular field of view, and improper calibration. These problems manifest themselves in cursor positioning error, wherein the cursor position is a small distance from the actual location of the user s gaze. This inherent error reduces the usability of the system, such that at the current time use with a generic computer desktop becomes impractical. The solution to this was to create an interface specialized for gaze-based interaction, increasing the size of screen objects to a size larger than the expected error of the system. RESULTS The designed system produces accurate tracking and cursor placement, as well as producing a reaction time superior to traditional tactile interfaces. Figure 1 shows an error contour plot depicting absolute error at specific screen locations. Low error is shown in blue and high error is shown in red Figure 1 goes here As shown in Figure 1, error is higher in the corner sections of the screen. The average error over the entire screen is 14 pixels. This resulted in a system that was very usable. Users demonstrated a strong capability to control the system. Along with accurately placing the cursor, the system also improved upon reaction and positioning times of standard computer input devices (such as a mouse or track pad.) Figure 2 shows average user response times for a track pad, a mouse, and the gaze based system Figure 2 goes here As shown, the gaze based system is inherently faster than other input methods. A 53% increase was seen over standard track pad methods, and a 12% increase over mouse based systems. Users claimed that the increase in speed was incredible, with the interface able to navigate to a page selected by the user almost as quickly as the user could think about navigating to that page. CONCLUSION

4 Page 3 of 7 RESNA The designed system has shown to be an acceptable low cost alternative to previous gaze tracking methods. REFERENCES (1) Duchowski, A A breadth-first survey of eye-tracking applications. Behavior Research Methods, Instruments and Computers, vol. 34, No. 4, pp (2) Jacob, R The use of eye movements in human-computer interaction techniques: what you look at is what you get. ACM Transactions on Information Systems, vol. 9, No. 2, pp (3) Majaranta, P., Raiha, K., Twenty years of eye typing: systems and design issues, Proceedings of the symposium on Eye tracking research and applications, 2002, pp (4) Hornof, A. J., Cavender, A., Hoselton, R., Eyedraw: A system for drawing pictures with eye movements, ACM SIGACCESS Conference on Computers and Accessibility, Atlanta, Georgia, 2004, pp (5) Silbert, L., Jacob, R., Evaluation of eye gaze interaction, Proceedings of the SIGCHI conference on Human factors in computing systems, 2000, pp (6) Tanriverdi, V., Jacob, R., Interacting with eye movements in virtual environments, Proceedings of the SIGCHI conference on Human factors in computing systems, 2000, pp (7) Young, L., Sheena, D., Survey of eye movement recording methods, Behavior Research Methods and Instrumentation 7, 1975, pp (8) Babcock J., Pelz, J., Building a lightweight eyetracking headgear, Eye Tracking Research and Applications Symposium, 2004, pp (9) Pelz, J., Canosa, R., Babcock J., Kucharczyk, D., Silver, A., Konno, D., Portable eyetracking: A study of natural eye movements Proceedings of the SPIE, Human Vision and Electronic Imaging, 2000, pp (10) Li, D., Babcock J., and Parkhurst, J.D.. openeyes: a low-cost head-mounted eye-tracking solution. AMC Eye Tracking Research and Applications Symposium, (11) Heidenburg, B., Lenisa, M., Wentzel, D., Malinowski, A., February 2008, Gaze Tracking System, Project Web Site, Available: (12) Heidenburg, B., Lenisa, M., Wentzel, D., Malinowski, A., Data Mining for Gaze Tracking System, To be published in Proceedings of Conference on Human System Interaction (HSI08), Krakow, Poland, May 25-27, CONTACT INFORMATION Michael J. Lenisa 750 Rosedale Roselle, IL CELL: (630) mlenisa@bradley.edu

5 RESNA Page 4 of 7 Figure 1 Figure 2

6 Page 5 of 7 RESNA!"#$#%&"''((%& )*"+($(%

7 RESNA Page 6 of 7 Shown is a three dimensional graph depicting error values for various cursor positions on the display.

8 Page 7 of 7 RESNA This chart shows a comparison of computer input methods and their respective reaction times. Input methods shown include: Track Pad (762.7 milliseconds), Standard mouse (559.9 milliseconds), and the designed gaze based system (496.9 milliseconds).

openeyes: a low-cost head-mounted eye-tracking solution

openeyes: a low-cost head-mounted eye-tracking solution Dongheng Li, Jason Babcock, and Derrick J. Parkhurst The Human Computer Interaction Program Iowa State University, Ames, Iowa, 50011 Abstract Eye