基于ASM和Gabor变换的人脸识别

人脸形状特征的变化是造成人物面貌差异的重要原因之一。主动形状模型(ASM)技术提供了检测该特征的有力手段。Gabor变换有着良好的仿生特性。它提供了理解视觉信息的有效途径。结合上述技术,采用点分布模型对人脸形状进行描述,利用ASM进行人脸特征点的搜索。以特征点上的Gabor展开系数作为人脸特征矢量,进行人脸辨识,实验结果表明,该算法能够在少量训练样本的情况下获得较高的识别率,并对光照、人物表情等变化具有较好的适应性。

ASM and gabor feature selection for face recognition

ASM is a useful method to detect geometric features. Gabor transform offers an efficient way of understanding visual information. A new face recognition method is presented based on ASM and gabor transform. PDM is used to describe face contour and ASM to search for face contour location. Face features are attained from gabor transform. Experiments show that this method achieves high recognition rate with few training samples and is robust to the variation of lighting and facial expressions.