融合局部纹理和形状特征的人脸表情识别

摘要

In order to improve the inadequacies of Local Binary Pattern (LBP), Center-Symmetric Local Binary Pattern (CS-LBP) and Histogram of Oriented Gradient (HOG) algorithm, Center-Symmetric Local Smooth Binary Pattern (CS-LSBP) and Histogram of Oriented Absolute Gradient (HOAG) are proposed, and a facial expression recognition method based on local texture and local shape features is proposed in this paper. Firstly, CS-LSBP and HOAG are used to extract two local features of expression image of the face. Then, Canonical Correlation Analysis (CCA) is used to fuse two local features. Finally, Support Vector Machine (SVM) is performed for the expression classification. Experimental results on JAFFE and Cohn-Kanade (CK) facial expression databases show that, the improved feature extraction method can extract the detail information of the image more completely and accurately. And the fusion method based on CCA can give full play to the representation ability of each feature. The facial expression recognition method proposed in this paper obtains a better recognition effect.%针对局部二值模式(LBP)、中心对称局部二值模式(CS-LBP)和梯度方向直方图(HOG)的不足进行改进,该文提出中心对称局部平滑二值模式(CS-LSBP)和绝对梯度方向直方图(HOAG),并提出一种融合局部纹理特征和局部形状特征的人脸表情识别方法.该方法首先采用CS-LSBP算子和HOAG算子分别提取人脸表情图像的局部纹理特征和局部形状特征,然后使用典型线性分析法(CCA)进行特征融合,最后利用支持向量机(SVM)进行表情分类.在JAFFE人脸表情库和Cohn-Kanade(CK)人脸表情库上的实验结果表明,改进的特征提取方法能更加完整、精确地提取图像的细节信息,基于CCA的特征融合方法能充分发挥特征的表征能力,该文所提人脸表情识别方法取得了较好的分类识别效果.