PROJECT PROPOSAL
EE 368 Digital Image Processing
May 2, 2000

* Title: Face Detection in Color Images

* Names:                             Email:

  Weng foo Henry Chang      henrycwf@stanford.edu
  Ulises Robles-Mellin           ulisesr@stanford.edu

* Description:

In this project we will present an algorithm and implementation of a
color based technique for detecting human faces in scenes (images)
where they appear in.

Some research has been done in this area. Usually, face detection is
achieved
by training neural networks and measuring distances between training
sets
in order to detect areas that might indicate a human face. Another
method
for doing face detection is by using grayscale and color information.
Therefore, we do not need to take the time, for instance, to train a
neural network. We will take the latter idea by implementing an
algorithm
to detect faces independently of the background color of the scene.

One way to approach this problem is by using a two step image processing
region
selection process. First, we will locate faces by searching for the skin

color. This can be done by looking for regions with the same RGB ratio
values that suggest human skin patterns. This part of the design is
important
since it will determine the final quality of the classification.

The face detection process follows. After computing the face region, we
convert the selected region to a grayscale image. We will then remove
zones that have the darkest (e.g. eyebrows) and brightest (e.g. lips)
gray level patterns. Facial features that have a different skin color
patterns
usually are: eyes, eyebrows, nostrils and mouth. The resulting image
will have regions with apertures or "holes". Finally, we will compare
the modified region that suggests a probable face to a pattern that
indicates a human face.

The project will be implemented in Matlab using the Matlab Image
Processing Toolkit.


* References:

[1] Yang J., Waibel A. A Real-Time Face Tracker, School of Computer
Science, Carnegie Mellon University.

[2] Menser B., Brunig M. Locating Human Faces in Color Images with
Complex Background

[3] Kapur J., Face Detection in Color Images. University of Washington.

[4] Menser B., Brunig M. Segmentation of Human Faces in Color Images
Using
Connected Operators