Face Detection In Color Images - PowerPoint PPT Presentation

About This Presentation
Title:

Face Detection In Color Images

Description:

Use YCbCr color space for good cluster separation. Model the skin and background color distributions with GMM ... Figure is obtain from Sung, Kah Kay (1996) ... – PowerPoint PPT presentation

Number of Views:1060
Avg rating:3.0/5.0
Slides: 15
Provided by: shaoh
Learn more at: https://web.stanford.edu
Category:
Tags: color | detection | face | images | kay

less

Transcript and Presenter's Notes

Title: Face Detection In Color Images


1
EE368 Project
Face Detection In Color Images
Wenmiao Lu Shaohua Sun
Group 3
2
EE368 Project
Skin Segmentation
  • Overview
  • Human Skin Segmentation
  • Adaptive Shape Analysis
  • View-based Face Detection
  • Results

Shape Analysis
Face Detection
3
EE368 Project
Human Skin Segmentation
  • Use YCbCr color space for good cluster separation
  • Model the skin and background color distributions
    with GMM
  • Segmentation by maximum likelihood classification

4
EE368 Project
An Example for Initial Skin Segmentation
Fairly complete skin segmentation with some noise
5
EE368 Project
Adaptive Shape Analysis
Refine the binary map
Open to get smaller regions
Initial Face Identification
Different Structuring Elements
Prior Information
Erosion Dilation
6
EE368 Project
An Example for Adaptive Shape Analysis
  • Medium size faces
  • Small, big or odd shaped regions passed to next
    stage

7
EE368 Project
View-Based Face Detection
Project to Low-dimensional Feature Space Spanned
by Largest Eigenvectors
Test Pattern
Face/Non-Face Decision
8
EE368 Project
Distances to Face Model
Test pattern is measured against the Face Model,
which consists of i) 6 Face Clusters and ii) 6
Non-face Clusters
Figure is obtain from Sung, Kah Kay
(1996)Learning and Example Selection for Object
and Pattern Detection.Ph.D. Thesis,
Massachusetts Institute of Technology, 1995.
9
EE368 Project
Distances between Test Pattern and One Cluster
Figure is obtain from Sung, Kah Kay
(1996)Learning and Example Selection for Object
and Pattern Detection.Ph.D. Thesis,
Massachusetts Institute of Technology, 1995.
10
EE368 Project
Neural Network Classification
  • 2-distance metric is discriminative for face and
    non-face patterns.
  • 2 distances have different magnitude neural
    network performs the final classification.

Figure is obtain from Sung, Kah Kay
(1996)Learning and Example Selection for Object
and Pattern Detection.Ph.D. Thesis,
Massachusetts Institute of Technology, 1995.
11
EE368 Project
Experimental Results
Detection Rate 95.6 False Positive 0.6
12
EE368 Project
13
(No Transcript)
14
(No Transcript)
Write a Comment
User Comments (0)
About PowerShow.com