Multimedia Security system project proposal

1
Multimedia Security system project proposal

• Face Detection and Tracking
• Present by Chun-Wei Hsu and Chih-Kai Yang

2
Introduction

• Face detection
• To verify which portion is a face in a image
• .
• Face Tracking
• After verifying a face, try to track the face
  motion.

3
Part 1 Overflow of Face Detection
4
Skin Color Filter

• Collect the enough skin database.
• Transform the RGB of face skin into YCbCr.
• Creating the threshold value to judge face skin.
• -Bayes Decision Rule

5
2.Face Template Research

• Creating the face templates from face database.
• To decide the candidate by comparing the input
  image with face template.

6
3. Verification

• To reduce the false detection.
• Face Region Analysis.

7
4.Eliminating Overlapping Detection

• Detected same face over one time.
• Eliminating the face candidate whose distance to
  corresponding template is large.

8
Part2. Face Tracking

• Tracking method
• a. Intensity gradient method.
• b. Color histogram method.

9
Intensity Gradient Method

• Model the head as an ellipse shape.
• Measure the match by the normalized sum of the
  gradient magnitude around the perimeter of the
  ellipse.

10
Color Histogram Method

• Find out the YCbCr histogram of the face.
• Tracking the location of face according to the
  histogram intersection.

11
Advantage of our methods

• Complementary.
• Robust.

12
Reference

• S.-D. Wei and S.-H. Lai, An Efficient Algorithm
  for Detecting Faces from Color Images, to appear
  in Proc. of Pacific Rim Conference on Multimedia,
  Hsinchu, Taiwan, Dec. 16-18, 2002.
• S. Birchfield. Elliptical head tracking using
  intensity gradients and color histograms. In
  Proceedings of the IEEE Conference on Computer
  Vision and Pattern Recognition, June 1998.
• D. Chai and A. Bouzerdoum ,A Bayesian Approach
  to Skin Color Classification in YCbCr Color
  Space IEEE Region Ten Conference (TENCON2000)
  ,Kuala Lumpur,Malaysia,vol.II,pp.421-424,Sep.2000.
  
• Douglas Chai , Son Lam Phung ,Abdesselan
  Bouzerdoum ,Skin Color Detection For Face
  Localization in Human-Machine Communications,
  IEEE International Symposium on Signal Processing
  Its Applications (ISSPA'2001), Kuala Lumpur,
  Aug. 2001.

13
Timeline