A Design of Face Detection and Facial Expression Recognition Techniques Based on Boosting Schema

Abstract:

Article Preview

During the development of the facial expression classification procedure, we evaluate three machine learning methods. We combine ABAs with CARTs, which selects weak classifiers and integrates them into a strong classifier automatically. We have presented a highly automatic facial expression recognition system in which a face detection procedure is first able to detect and locate human faces in image sequences acquired in real environments. We need not label or choose characteristic blocks in advance. In the face detection procedure, some geometrical properties are applied to eliminate the skin color regions that do not belong to human faces. In the facial feature extraction procedure, we only perform both the binarization and edge detection operations on the proper ranges of eyes, mouth, and eyebrows to obtain the 16 landmarks of a human face to further produce 16 characteristic distances which represent a kind of expressions. We realize a facial expression classification procedure by employing an ABA to recognize six kinds of expressions. The performance of the system is very satisfactory; whose recognition rate achieves more than 90%.

Info:

Periodical:

Edited by:

Dongye Sun, Wen-Pei Sung and Ran Chen

Pages:

617-621

DOI:

10.4028/www.scientific.net/AMM.121-126.617

Citation:

C. Y. Kao and C. S. Fahn, "A Design of Face Detection and Facial Expression Recognition Techniques Based on Boosting Schema", Applied Mechanics and Materials, Vols. 121-126, pp. 617-621, 2012

Online since:

October 2011

Export:

Price:

$35.00

In order to see related information, you need to Login.

In order to see related information, you need to Login.