基于POEM_SLPP的人臉識(shí)別算法
發(fā)布時(shí)間：2018-10-08 15:35
【摘要】：針對(duì)方向邊緣幅值模式(patterns of oriented edge magnitudes,POEM)提取的人臉特征維數(shù)過高和計(jì)算復(fù)雜度較大的問題,提出了結(jié)合方向邊緣幅值模式和有監(jiān)督的局部保持投影(patterns of oriented edge magnitudes_supervised locality preserving projections,POEM_SLPP)的人臉識(shí)別算法。首先,采用POEM算子進(jìn)行特征提取;其次,將高維特征數(shù)據(jù)投影到SLPP算法求出的低維樣本空間進(jìn)行降維;最后,采用最近鄰法對(duì)測(cè)試樣本進(jìn)行分類。在CAS-PEAL-R1人臉庫(kù)上的實(shí)驗(yàn)結(jié)果表明,在表情、背景、飾物、時(shí)間、距離測(cè)試集上,該算法的平均識(shí)別率較POEM+LPP算法提高了22%,較POEM+PCA提高了2%。
[Abstract]:The dimension of face feature extracted by directional edge amplitude mode (patterns of oriented edge magnitudes,POEM) is too high and the computational complexity is large. A face recognition algorithm based on directional edge amplitude pattern and supervised local preserving projection (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 into the low-dimensional sample space of SLPP algorithm for dimensionality reduction; finally, the nearest neighbor method is used to classify the test samples. The experimental results on CAS-PEAL-R1 face database show that the average recognition rate of the algorithm is 22% higher than that of POEM LPP algorithm and 2% higher than that of POEM PCA in expression, background, ornaments, time and distance test sets.
【作者單位】： 上海電力學(xué)院自動(dòng)化工程學(xué)院;
【基金】：上海市電站自動(dòng)化技術(shù)重點(diǎn)實(shí)驗(yàn)室資助項(xiàng)目(13DZ2273800)
【分類號(hào)】：TP391.41
[Abstract]:The dimension of face feature extracted by directional edge amplitude mode (patterns of oriented edge magnitudes,POEM) is too high and the computational complexity is large. A face recognition algorithm based on directional edge amplitude pattern and supervised local preserving projection (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 into the low-dimensional sample space of SLPP algorithm for dimensionality reduction; finally, the nearest neighbor method is used to classify the test samples. The experimental results on CAS-PEAL-R1 face database show that the average recognition rate of the algorithm is 22% higher than that of POEM LPP algorithm and 2% higher than that of POEM PCA in expression, background, ornaments, time and distance test sets.
【作者單位】： 上海電力學(xué)院自動(dòng)化工程學(xué)院;
【基金】：上海市電站自動(dòng)化技術(shù)重點(diǎn)實(shí)驗(yàn)室資助項(xiàng)目(13DZ2273800)
【分類號(hào)】：TP391.41
【相似文獻(xiàn)】
相關(guān)期刊論文 前10條
1 王李冬;;一種新的人臉識(shí)別算法[J];計(jì)算機(jī)技術(shù)與發(fā)展;2009年05期
2 桂榮;沈榮鑫;詹泳;;基于模糊神經(jīng)推理的人臉識(shí)別算法研究[J];華南師范大學(xué)學(xué)報(bào)(自然科學(xué)版);2012年04期
3 蘇士美;王燕;王明霞;;基于加權(quán)小波分解的人臉識(shí)別算法研究[J];鄭州大學(xué)學(xué)報(bào)(工學(xué)版);2014年01期
4 魏冬梅;周衛(wèi)東;;采用壓縮感知的人臉識(shí)別算法[J];計(jì)算機(jī)工程;2011年18期
5 陳皓;霍星;;視頻監(jiān)控中人臉識(shí)別算法穩(wěn)定性的改進(jìn)[J];工程圖學(xué)學(xué)報(bào);2011年06期
6 李美瑾;;人臉識(shí)別算法綜述與遠(yuǎn)程考試系統(tǒng)應(yīng)用研究[J];廣西廣播電視大學(xué)學(xué)報(bào);2012年01期
7 盛朝強(qiáng);王君;;煤礦井下人員簽到系統(tǒng)人臉識(shí)別算法研究[J];計(jì)算機(jī)技術(shù)與發(fā)展;2012年07期
8 朱學(xué)毅;王崇駿;周新民;張W，
本文編號(hào)：2257303
本文鏈接：http://sikaile.net/kejilunwen/ruanjiangongchenglunwen/2257303.html
最近更新
教材專著
