摘要
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,在表情、背景、饰物、时间、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%。
Considering that facial feature extracted by the patterns of oriented edge magnitudes had the high dimensionality and complex computing, this paper proposed a face recognition algorithm based on the patterns of oriented edge magnitudes_ super- vised locality preserving projections（POEM_SLPP）. This algorithm first extracted facial feature by POEM operator, and then got dimension reduction by projecting the high-dimensional feature data to the sample space obtained by SLPP algorithm. Finally, the proposed algorithm classified test samples by nearest neighbor method. Experimental results on CAS-PEAL-R1 face database （including the expression, background, accessory, age, distance test set） indicate that the average recognition rate of the new algorithm increases by 22% than the POEM ＋ LPP algorithm, and increases by 2% than the POEM ＋ PCA algorithm.
出处
《计算机应用研究》
CSCD
北大核心
2017年第6期1896-1899,共4页
Application Research of Computers
基金
上海市电站自动化技术重点实验室资助项目(13DZ2273800)
关键词
人脸识别
方向边缘幅值模式
有监督的局部保持投影
face recognition
patterns of oriented edge magnitudes （ POEM ）
supervised locality preserving projections （SLPP）
