人脸识别系统在嵌入式ARM上的实现

由于通用RU-LBP和PCA结合的人脸识别算法得到的子空间内各类之间距离不均衡导致人脸识别低，提出一种岭回归人脸识别方法。通过使用 RU-LBP 和 PCA 进行人脸特征提取，然后对这些特征使用岭回归算法进行识别，该算法能够最大程度上均化各类之间的距离，提高分类能力。在 FERET 标准的人脸库上的实验结果表明，该算法能够显著改进人脸识别率。

Implementation of face recognition system based on embedded ARM

Due to most face recognition algorithms based on PCA and general RU-LBP get a low recognition rate result from distance imbalance in subspace, a face recognition method called ridge regression is proposed in this paper. Through face feature extraction by RU-LBP and PCA, apply the ridge regression algorithm to face recognition. It maximizes margins between the data of different classes to improve the objective function. As the experimental results on FERET standard face database shown, the method achieves better performance than the common algorithm.