Abstract: To overcome the limitations of traditional face recognition methods for single sample under variations in position, expression, occlusion and illumination, an improved face descriptor ε-WLBD （ε-Weber Local Binary Descriptor） based on texture features and gradient information extraction is proposed. Firstly, improved Local Binary Pattern（LBP） is used for texture features extraction and improved Kirsch for edge features extraction. Then the histogram statistics are conducted respectively and concatenated into the general feature vector. The nearest neighbor classifier is used for face image classification and recognition. Compared with a variety of algorithms, experimental results on YALE face database and AR face database indicate that the proposed method is simple and effective, and robust to variations of face position, expression, occlusion and illumination, and also has better performance in face description for single sample.

Key words: face recognition, ε-Weber Local Binary Descriptor（ε-WLBD）, Local Binary Pattern（LBP）, Kirsch, single sample

摘要： 针对传统人脸识别方法在单样本条件下受姿态、表情、遮挡和光照影响识别效果不佳等问题，提出一种改进的纹理特征和边缘特征相结合的人脸描述算子ε-WLBD（ε-Weber Local Binary Descriptor）。先用改进的局部二值模式和改进的Kirsch算子进行纹理特征和边缘特征提取，然后分别进行直方图统计，并将其串接起来作为人脸识别的总体特征向量，最后利用最近邻算法进行分类识别。在YALE和AR人脸库上进行测试，实验结果表明所提方法简单有效，且对姿态、表情、遮挡和光照等变化具有较强鲁棒性，对单样本人脸描述具有较好的效果。

关键词: 人脸识别, &epsilon, -WLBD算子, 局部二值模式, Kirsch算子, 单样本