Face Recognition Algorithm Based on Neighborhood Preserving Discriminant Embedding

摘要：

　　Neighborhood Preserving Embedding (NPE) is a linear algorithm which can preserve the local geometric structure information of data after mapping however, as an unsupervised algorithm, the extracted features for classification are not optimal.In this paper, we propose a supervised algorithm for face recognition which is named Neighbourhood Preserving Discriminating Embedding algorithm (NPDE).Firstly, the algorithm calculates the between-class scatter matrix and within-class scatter matrix of sample datas, at the same time, introduces the subtraction of the between-class scatter matrix and within-class scatter matrix into the objective function of the NPE algorithm, so the NPDE algorithm can preserve the local geometric structure and more discriminative as well.The experimental results on ORL and YALE databases demonstrate that the recognition rate of NPDE algorithm for face recognition is significantly increased.

关键词： Neighborhood Preserving Embedding Maximum Margin Criterion Face Recognition

作者: Yan WANG Huali LIU Wenjun SUN

作者单位: College of Computer and Communication,Lanzhou University of Technology,Lanzhou,China

会议类型: 国内会议

会议名称: 第四届信息电子与计算机工程国际会议(The 4th International Conference on Information ,Electronic and Computer Science)

会议地点: 泰安

会议语种:英文

页码: 123-127

在线出版日期: 2012-11-24（万方平台首次上网日期，不代表论文的发表时间）