基于POEM_SLPP的人脸识别算法

针对方向边缘幅值模式(Patterns of Oriented Edge Magnitudes,POEM)提取的人脸特征维数过高和计算复杂度较大的问题,提出了结合方向边缘幅值模式和有监督的局部保持投影(Patterns of Oriented Edge Magnitudes _Supervised Locality Preserving Projections,POEM_SLPP)的人脸识别算法。首先,采用POEM算子进行特征提取；其次,将高维特征数据投影到SLPP算法求出的低维样本空间进行降维；最后,采用最近邻法对测试样本进行分类。在CAS-PEAL-R1人脸库上的实验结果表明,在姿态、背景、修饰、年龄、距离测试集上,该算法的平均识别率较POEM LPP算法提高了22%,较POEM PCA提高了2%。

Face recognition algorithm based POEM_ SLPP

Since facial feature extracted by the patterns of oriented edge magnitudes has the high dimensionality and complex computing, face recognition algorithm based on the patterns of oriented edge magnitudes_ supervised locality preserving projections(POEM_SLPP) is proposed. Firstly, POEM operator is used for feature extraction; Secondly, the high-dimensional feature data is projected to the sample space obtained by SLPP algorithm; Finally, test samples classified by nearest neighbor method. Experimental results on CAS-PEAL-R1 face database indicate that the average recognition rate of the new algorithm increases by 22% than the POEM LPP algorithm, increased by 2% than the POEM PCA algorithm on the pose, background, accessory, age, distance test set.