基于流形学习和正交稀疏保留投影的人脸和掌纹图像特征提取方法
首发时间:2013-03-12
摘要:对于一个数据集,数据间的稀疏重构关系具有很好的分类信息。稀疏保留投影(SPP)正是基于这样的考虑所提出的一种特征提取方法,它的目标是获取一个线性投影空间,使得样本之间的全局稀疏重构关系得以保留。然而,稀疏保留投影主要关注的是样本间的全局稀疏重构关系,并且得到的投影变换通常不是正交的,而在实际应用中,图像数据往往处于高维空间中的一种低维流形结构中,正交性一直被认为有利于提高鉴别能力。因此,本文在稀疏保留投影中引入了流形结构保留和正交投影,提出了两种实现算法来实现人脸和掌纹图像的特征提取,分别是基于流形学习的整体正交稀疏保留投影(MLHOSPP)和基于流形学习的迭代正交稀疏保留投影(MLIOSPP)。
关键词: 模式识别 特征提取 流形学习 子空间学习 正交稀疏保留投影 人脸和掌纹图像
For information in English, please click here
Face and Palmprint Image Feature Extraction based on Manifold Learning and Orthogonal Sparsity Preserving Projections
Abstract:For a given data set, global sparse reconstruction relations among its data has been shown to contain useful information for classification. Based on this, a feature extraction method named sparsity preserving projections (SPP) has been proposed, which tends to seek a linear projected subspace where the global sparse reconstruction relations among samples could be preserved. However, SPP focuses on the global sparse reconstruction relations among samples, and its achieved transformation usually is not orthogonal. In real world, image samples possibly reside on a nonlinear submanifold of the high-dimensional space, which is the inherent structure among the samples, and orthogonality is favorable for classification in many scenarios. Therefore, in this paper, we introduce manifold preserving and orthogonal transformation into SPP, and propose two novel approaches for face and palmprint image feature extraction, which are manifold learning based holistic orthogonal sparsity preserving projections (MLHOSPP) and manifold learning based iterative orthogonal sparsity preserving projections (MLIOSPP).?
Keywords: pattern recognition feature extraction manifold learning subspace learning orthogonal sparsity preserving projections face and palmprint images
论文图表:
引用
No.****
同行评议
共计0人参与
勘误表
基于流形学习和正交稀疏保留投影的人脸和掌纹图像特征提取方法
评论
全部评论0/1000