摘要
针对压缩感知理论中的核心问题,即如何通过有限的测量值以较高的重建率重构稀疏信号,提出了基于主元分析和压缩感知的人脸识别方法(PSL0).该算法利用双向二维主成分分析提取图像行列2个方向的特征并进行降维,建立反映人脸特征投影矩阵,作为压缩感知算法的超完备基,将每一幅待识别图像的特征向量作为测量值,用基于平滑l0范数快速稀疏表示(SL0)算法求解l0范数最小化问题,寻求图像在该超完备基上的稀疏表示,以得到一组最优稀疏系数重构各类图像,求取测试图像与各类重构图像的最小残差进行分类识别.实验结果表明,该算法在同类算法中获得了较高的人脸识别率及较好的重建效果.
A face recognition method based on Compressed Sensing(CS) algorithm on principal component analysis is proposed according to the core problem of the theory of compressed sensing,such as how to reconstruct sparse signal from limited measurements.It utilizes the two directional PCA((2D)2PCA) transform to extract images features in both row and colum directions and reduce the dimension.A projection matrix is constructed to identify the face features,considering these features to form an over complete dictionary and the feature vector of each test images is the observation value.By solving the l0 norm minimization,it finds out the sparsest representation of images based on the dictionary to obtain a set of optimal sparse coefficient,which is used to recover the train images,computes the residuals between test and train images for face recognition.Experimental results show that this method not only has a high recognition rate in a lower dimension,but also reduces the computational complexity.
出处
《西安工程大学学报》
CAS
2013年第4期524-529,共6页
Journal of Xi’an Polytechnic University
关键词
人脸识别
压缩感知
稀疏表示
最小l0范数
face recognition
compressed sensing(CS)
sparse representation
minimization l0 norm