Eye Detection
After detecting the face, the ey
extracted. Eyes are the most important
Since, eyes lie in the upper half of the
face is removed in order to reduce s
many applications of the robust eye sta
example, the eye states provide impo
recognizing facial expression and hum
systems. In this paper, a method is pres
states in nearly frontal-view color fac
using the Hough transform. The eye i
and important feature in a human fac
eyes are often easier as compared to oth
detection is done by using various me
which is used in this paper for detectio
Hough transform. After detecting the
frame next step is to extract the eyes.
search area a dummy image is create
mask over the image. And then fill t
using imfill command which fills the
image. In order to get the eye region
multiply the dummy image with that
image, whose output becomes the are
eye part. Result of this are as follows.
Figure.5 Extraction of ey
The general Hough transform can b
shape, although the complexity of the t
with the number of parameters needed
The Hough transform is a method that,
to find features of any shape in an im
only generally used for finding straigh
The parameter space is defined accord
object of interest. A straight line pass
(x1, y1) and (x2, y2) can in the x, y pla
y = ax+b
The Hough transform for lines
representation of lines, since lines perp
will have a value of infinity. This wi
space a, b to have infinite size. Instea
ye regions are to be
feature of human face.
face, the lower part of
search area. There are
ates extraction [8]. For
ortant information for
man-computer interface
sented to detect the eye
ce image sequence by
is the most significant
e, as extraction of the
her facial features. Eye
thods [9]. The method
on of eye is the circular
face from input video
In order to reduce the
ed, which is used as a
this dummy image by
e holes in the dummy
n from the search area
t of the face detected
ea which contains the