By Topic

Applying Contrast-limited Adaptive Histogram Equalization and integral projection for facial feature enhancement and detection

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

3 Author(s)
Yisu Zhao ; Distributed and Collaborative Virtual Environments Research Laboratory (DISCOVER), School of Information Technology and Engineering, University of Ottawa, K1N 6N5, Canada ; Nicolas D. Georganas ; Emil M. Petriu

In order to achieve real-time subject-independent automatic facial feature enhancement and detection, a novel method is presented in this paper combing Contrast-limited Adaptive Histogram Equalization (CLAHE) and multi-step integral projection. First, after real-time detecting face images, a sigma filter is used to remove the noise in images. Sigma filtering is chosen in this research because of its validity in noise removal. It has the advantages of providing a good noise removal result, not blurring the image and fast performance. Second, since it is important to extract facial features as accurately and clearly as possible, CLAHE is then applied on images for enhancing the facial features. This step is done after the sigma filter in order not to amplify the noise in images. Third, after enhancing these features, multi-step integral projection is proposed to detect the useful facial features regions automatically. Finally, the detected facial feature region is then extracted by Gabor transformation and the final facial expression recognition is classified by SVMs. We test our system on the JAFFE database and achieve a high recognition rate of 95.318% on trained data.

Published in:

Instrumentation and Measurement Technology Conference (I2MTC), 2010 IEEE

Date of Conference:

3-6 May 2010