發(fā)布時間：2018-03-21 05:15

  本文選題：低秩類字典　切入點：稀疏誤差字典　出處：《中國圖象圖形學(xué)報》2017年09期 　論文類型：期刊論文

【摘要】：目的由于受到光照變化、表情變化以及遮擋的影響,使得采集的不同人的人臉圖像具有相似性,從而給人臉識別帶來巨大的挑戰(zhàn),如果每一類人有足夠多的訓(xùn)練樣本,利用基于稀疏表示的分類算法(SRC)就能夠取得很好地識別效果。然而,實際應(yīng)用中往往無法得到尺寸大以及足夠多的人臉圖像作為訓(xùn)練樣本。為了解決上述問題,根據(jù)基于稀疏表示理論,提出了一種基于聯(lián)合判別性低秩類字典以及稀疏誤差字典的人臉識別算法。每一類的低秩字典捕捉這類的判別性特征,稀疏誤差字典反映了類變化,比如光照、表情變化。方法首先利用低秩分解理論得到初始化的低秩字典以及稀疏字典,然后結(jié)合低秩分解和結(jié)構(gòu)不相干的理論,訓(xùn)練出判別性低秩類字典和稀疏誤差字典,并把它們聯(lián)合起來作為測試時所用的字典;本文的方法去除了訓(xùn)練樣本的噪聲,并在此基礎(chǔ)上增加了低秩字典之間的不相關(guān)性,能夠提高的低秩字典的判別性。再運(yùn)用l1范數(shù)法(同倫法)求得稀疏系數(shù),并根據(jù)重構(gòu)誤差進(jìn)行分類。結(jié)果針對Extended Yale B庫和AR庫進(jìn)行了實驗。為了減少算法執(zhí)行時間,對于訓(xùn)練樣本利用隨機(jī)矩陣進(jìn)行降維。本文算法在Extended Yale B庫的504維每類32樣本訓(xùn)練的識別結(jié)果為96.9%。在無遮擋的540維每類4樣本訓(xùn)練的AR庫的實驗結(jié)果為83.3%,1 760維的結(jié)果為87.6%。有遮擋的540維每類8樣本訓(xùn)練的AR庫的結(jié)果為94.1%,1 760維的結(jié)果為94.8%。實驗結(jié)果表明,本文算法的結(jié)果比SRC、DKSVD(Discriminative K-SVD)、LRSI(Low rank matrix decomposition with structural incoherence)、LRSE+SC(Low rank and sparse error matrix+sparse coding)這4種算法中識別率最高的算法還要好,特別在訓(xùn)練樣本比較少的情況下。結(jié)論本文所提出的人臉識別算法具有一定的魯棒性和有效性,尤其在訓(xùn)練樣本較少以及干擾較大的情況下,能夠取得很好地識別效果,適合在實際中進(jìn)行應(yīng)用。
[Abstract]:Objective because of the influence of illumination change, facial expression change and occlusion, the human face images collected from different people are similar, which brings a great challenge to face recognition, if there are enough training samples for each kind of people. The classification algorithm based on sparse representation can achieve good recognition results. However, in practical applications, face images with large size and enough face images can not be used as training samples. Based on sparse representation theory, a face recognition algorithm based on joint discriminant low rank dictionaries and sparse error dictionaries is proposed. Methods the initialized low rank dictionary and sparse dictionary are obtained by using the theory of low rank decomposition, and then the discriminant low rank dictionary and sparse error dictionary are trained by combining the theory of low rank decomposition and structural incoherence. The method in this paper removes the noise of the training sample and increases the non-correlation between the low-rank dictionaries. The sparse coefficient is obtained by using l 1 norm method (homotopy method) and classified according to the reconstruction error. Results the experiments are carried out on Extended Yale B library and AR library. In order to reduce the execution time of the algorithm, For training samples, random matrix is used for dimensionality reduction. The recognition result of 32 samples per class of 504 dimension in Extended Yale B library is 96.9. The experimental result of AR library with 4 samples training in 540D without occlusion is 83.3 dimensional. The result is 87.6. The AR library with 540 dimensions of occlusion and 8 kinds of samples trained is 94.1D and 1,760D, 94.80.The experimental results show that, The result of this algorithm is better than that of the four algorithms (SRC DKSVD discriminative K-SVD and LRSI low rank matrix decomposition with structural incorencein LRSE SC(Low rank and sparse error matrix sparse). Conclusion the face recognition algorithm proposed in this paper is robust and effective, especially in the case of less training samples and more interference. Suitable for practical application.
【作者單位】： 南京航空航天大學(xué)自動化學(xué)院;
【基金】：國家自然科學(xué)基金項目(61473148,U1531110) 江蘇省普通高校專業(yè)學(xué)位研究生創(chuàng)新計劃(SJLX16_0106)~~
【分類號】：TP391.41

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 延秀娟;陳永鋒;;人臉識別算法綜述[J];商場現(xiàn)代化;2008年26期

2 王李冬;;一種新的人臉識別算法[J];計算機(jī)技術(shù)與發(fā)展;2009年05期

3 桂榮;沈榮鑫;詹泳;;基于模糊神經(jīng)推理的人臉識別算法研究[J];華南師范大學(xué)學(xué)報(自然科學(xué)版);2012年04期

4 蘇士美;王燕;王明霞;;基于加權(quán)小波分解的人臉識別算法研究[J];鄭州大學(xué)學(xué)報(工學(xué)版);2014年01期

5 魏冬梅;周衛(wèi)東;;采用壓縮感知的人臉識別算法[J];計算機(jī)工程;2011年18期

6 陳皓;霍星;;視頻監(jiān)控中人臉識別算法穩(wěn)定性的改進(jìn)[J];工程圖學(xué)學(xué)報;2011年06期

7 李美瑾;;人臉識別算法綜述與遠(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;2012年01期

8 盛朝強(qiáng);王君;;煤礦井下人員簽到系統(tǒng)人臉識別算法研究[J];計算機(jī)技術(shù)與發(fā)展;2012年07期

9 朱學(xué)毅;王崇駿;周新民;張W，

本文編號：1642406