Академический Документы
Профессиональный Документы
Культура Документы
Technology
1
1
2
Yih-Ran Sheu
Juing-Shian Chiou
Cheng-Hsin Chuang
1
1
1
Nguyen Van Binh
Wen-Li Shen
Yao-Wei Tseng
1
Department of Electrical Engineering,
2
Department of Mechanical Engineering,
Southern Taiwan University of Science and Technology,
Tainan, TAIWAN
sheu@mail.stust.edu.tw
Abstract
The objective of this paper present a set of techniques
integrated into an eye-mouse system using a field
programmable gate array (FPGA). Our system requires a
camera (TRDB-D5M), FPGA-DE2 board and a personal
computer. A field programmable gate array (FPGA) is used
to control the CCD camera and execute the image
processing operations and control eye-mouse of computer
via RS232 protocol. The coordinate of eye-mouse in our
system is determined by the system based on FPGA
imaging analysis. No computer is necessary in image
processing. Good efficiency can be seen using FPGA
hardware approach to determine the eye-mouse coordinate
per image frame. The color information of eye is used to
estimate the direction of pupil movement and then uses the
direction information to control the mouse of computer.
Keywords: FPGA, eye mouse, direction of pupil
I. INTRODUCTION
Nowadays, eye tracking and eye mouse system more and
more are widely. There are many eye tracking system using
image processing by computer supported by many image
processing tools such as Matlab, Visual C, Visual
Basicetc, but these systems reduce the speed of the
computer and sometimes cause unexpected situations. Thus,
processing image to detect eye movements using FPGA
hardware bring an alternative solution. In this paper, we
practically design and develop the improved eye-mouse
device with TRD-D5M camera, FPGA hardware, and
computer. An eye tracking systems is a system that can
track the movements of the user's eyes. Eye mouse system
is a very good approach for people who do not have the
ability to perform the movements of common control on
computer such as mouse drag and mouse clicketc. In this
paper, we perform image processing to detect pupil
movement in FPGA hardware then sends control
information to the computer to control a cursor on the
screen. There are several different ways to track the
directions of eye movements. Each approach has its own
advantages, disadvantages, and limitations. We use
information about eye color to detect the movement of the
eye. There are many color spaces such as RGB, HSI, YUV,
YIQ ... We choose the YUV color space because it has
many advantages and simple. The image processing in
FPGA is not as good as image processing in personal
computer but this process is totally independent, thus
a. The definition of
thresholding
g(i,j) =
255, f(i, j) T
(2)
0, f(i, j) T
a. Original image
- 0.148 -
0.289
0.437 G (1)
a. Y image
1 K
1 K
XC,YC K Y (3)
K 1
1
X,
ATTENTION!
TRIAL LIMITATION - ONLY 3 SELECTED PAGES MAY BE CONVERTED PER CONVERSION.
Where K is the total number of the dark spot at which the
gray level is lower than the threshold after dynamic
binarizing of the eye image. X and Y are coordinates of the
black pixel. [5]
PURCHASING A LICENSE REMOVES THIS LIMITATION. TO DO SO, PLEASE CLICK ON THE FOLLOWING LINK:
http://www.investintech.com/order_a2e_pro.htm