中文摘要：

提出一种加权的局部二值模式和块线性判别投影的人脸识别方法，它首先编码人脸局部区域中心像素与邻域像素的灰度差的符号分量，体现了人脸局部结构的重要性．其次利用了局部二值模式的幅值分量作为像素局部二值模式的权重，最后利用块线性判别投影降低提出的描述符的特征维数，同时增强它的判别能力．在FERET人脸库的大量实验结果表明该算法可以获得有效的性能提升．

英文摘要：

It presents a weighted local binary pattern and block-based linear discriminant pro- jection for face recognition, which first encodes the sign components of gray difference bewteen cen- ter pixel and its neighbors, and express the importance of local face structure. Secondly, it uses the magnitude components as the weight of local binary pattern. Finally, it applies block-based lin- ear discriminant projection to reduce dimension of the propose descriptor and at the same time en- hance its discriminative power. Experimental results on FERET face database show that the pro- posed method can get effective performance improvement.