自适应加权完全局部二值模式的表情识别

为了有效地提取局部特征和全局特征以提高表情识别的性能，提出自适应加权的完全局部二值模式（Adaptively Weighted Compound Local Binary Pattern，AWCLBP）的人脸表情识别算法。首先对人脸表情图像进行预处理分离出表情子区域，与此同时生成表情子区域的贡献度图谱（Contribution Map，CM）；然后对表情子区域和整幅表情图像做完全局部二值模式变换提取三种特征（差值符号特征CLBP_S、差值幅值特征CLBP_M、中心像素特征CLBP_C）并连接三种特征生成级联直方图，并根据CM对表情子区域的级联直方图进行加权和整张图像的直方图进行融合；最后用卡方距离和最近邻方法进行分类识别。本算法在JAFFE库上做了实验并和LBP、Gabor小波、活动外观模型进行了比较，验证了本算法的有效性。

Facial expression recognition based on AWCLBP

In order to improve the performance of facial expression recognition by extracting the effective local features and global features, an algorithm that facial expression recognition based on Adaptively Weighted Compound Local Binary Pattern is proposed. Firstly, the facial expression sub-regions are isolated by preprocessing, and the Contribution Maps(CM) of facial expression sub-regions are computed; Secondly, the three features of expression sub-regions and the global facial expression image are extracted by Compound Local Binary Pattern, then cascade histograms are generated by connecting the histograms of the three features of the image and the l expression sub-regions which are weighted according to the CMs. Finally, the weighted cascade histograms are classified and recognized by using the chi-square distance and nearest neighbor method. Experiment results on the facial expression database of JAFFE show that the proposed algorithm can be applied to achieve the higher recognition rate than other algorithms, Such as, LBP, Gabor Wavelet and Active appearance model.