基于虛擬樣本的改進人臉識別算法研究及應用
發(fā)布時間:2018-12-14 12:33
【摘要】:針對實際采集的視頻中背景復雜,人物多變,圖像處理時間長,訓練樣本不足的問題,提出了構造虛擬樣本,并結合Gabor濾波器及對PCA-LDA算法加以改進的人臉識別算法,以應用于教室點名系統(tǒng)。首先對教室采集到的視頻進行裁剪,按幀截取并檢測出含有人臉的部分圖像并單獨保存為測試圖像,然后將其與已有人臉庫里的訓練圖像進行對比,最后采用提出的鏡像法構造虛擬樣本,并結合了Gabor濾波器以及PCNN灰度圖像增強處理算法的改進PCA-LDA算法進行人臉識別。仿真實驗表明,提出的算法預測了樣本可能存在的變化,也在一定程度上降低了計算復雜度,明顯地提高了識別率,并在教室點名系統(tǒng)中得到了較好的驗證。
[Abstract]:Aiming at the problems of complex background, changeable characters, long image processing time and insufficient training samples, a new face recognition algorithm based on Gabor filter and PCA-LDA algorithm is proposed. To apply to the classroom roll call system. First of all, the video collected in the classroom is clipped, the part of the face image is captured and detected according to the frame, and stored separately as the test image, and then it is compared with the training image in the existing human face database. Finally, the proposed image method is used to construct the virtual sample, and the improved PCA-LDA algorithm, which combines the Gabor filter and the PCNN gray image enhancement algorithm, is used for face recognition. The simulation results show that the proposed algorithm can predict the possible changes of the samples, reduce the computational complexity to a certain extent, and improve the recognition rate obviously, and is well verified in the classroom-roll call system.
【作者單位】: 江南大學物聯(lián)網(wǎng)工程學院;
【基金】:國家自然科學基金(No.61672265) 江蘇省產(chǎn)學研聯(lián)合創(chuàng)新資金項目(No.BY2013015-35)
【分類號】:TP391.41
[Abstract]:Aiming at the problems of complex background, changeable characters, long image processing time and insufficient training samples, a new face recognition algorithm based on Gabor filter and PCA-LDA algorithm is proposed. To apply to the classroom roll call system. First of all, the video collected in the classroom is clipped, the part of the face image is captured and detected according to the frame, and stored separately as the test image, and then it is compared with the training image in the existing human face database. Finally, the proposed image method is used to construct the virtual sample, and the improved PCA-LDA algorithm, which combines the Gabor filter and the PCNN gray image enhancement algorithm, is used for face recognition. The simulation results show that the proposed algorithm can predict the possible changes of the samples, reduce the computational complexity to a certain extent, and improve the recognition rate obviously, and is well verified in the classroom-roll call system.
【作者單位】: 江南大學物聯(lián)網(wǎng)工程學院;
【基金】:國家自然科學基金(No.61672265) 江蘇省產(chǎn)學研聯(lián)合創(chuàng)新資金項目(No.BY2013015-35)
【分類號】:TP391.41
【相似文獻】
相關期刊論文 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];計算機技術與發(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];華南師范大學學報(自然科學版);2012年04期
4 蘇士美;王燕;王明霞;;基于加權小波分解的人臉識別算法研究[J];鄭州大學學報(工學版);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[J];工程圖學學報;2011年06期
7 李美瑾;;人臉識別算法綜述與遠程考試系統(tǒng)應用研究[J];廣西廣播電視大學學報;2012年01期
8 盛朝強;王君;;煤礦井下人員簽到系統(tǒng)人臉識別算法研究[J];計算機技術與發(fā)展;2012年07期
9 朱學毅;王崇駿;周新民;張W,
本文編號:2378620
本文鏈接:http://sikaile.net/kejilunwen/ruanjiangongchenglunwen/2378620.html
最近更新
教材專著