Facial Expression Recognition Based on Salient Local Texture-based Features

• Main Authors: Hung-chiang Hsieh, 謝鴻瑲
• Other Authors: Chih-chin Lai
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/51647322680031512087

碩士 === 國立高雄大學 === 電機工程學系碩士班 === 102 === Automatic facial expression recognition has gained an increasing interest in recent years in building natural human-computer interaction systems. Extracting the discriminative features from facial images is the most important part of facial expression recognition. In this paper, we propose a facial expression recognition approach based on salient local texture-based features. We use Speeded Up Robust Features to find interest points and to generate salient regions from facial images, and then the associated features based on Center-Symmetric Local Ternary Pattern and Local Sign Directional Pattern in these regions are extracted. These features are classified by using the Support Vector Machine. Experimental results show that our approach can achieve higher recognition rate on the Cohn-Kanade, Extended Cohn-Kanade, TFEID, and JAFFE facial expression databases.