基于二维Gabor小波与AR-LGC的人脸特征提取算法研究
|
摘要
针对Gabor小波提取的特征信息不够全面以及维数灾难问题,提出了一种基于二维Gabor小波与AR-LGC人脸特征提取的改进算法。首先利用二维Gabor小波提取归一化人脸图像的特征信息,使用AR-LGC算子对提取的Gabor特征进行编码;然后对编码后的Gabor特征图谱进行划分,对每个子块进行直方图统计,将其串联形成人脸表情特征向量并利用PCA进行数据压缩,最后利用C-SVM进行分类识别。实验结果表明,平均识别率为93.33%的比AR-LGC+SVM,Gabor+PCA+SVM提取特征算法的平均识别率分别高6.19%,3.33%。因此所提出的算法在人脸图像的特征信息提取方面有一定的参考价值。
Aiming at the incompleteness problem that the feature information extracted from Gabor wavelet and the dimension of disaster, an improved algorithm with face feature extraction based on two-dimensional Gabor wavelet and AR-LGC is proposed. In this algorithm, normalized face features are extracted by two-dimensional Gabor wavelet and the extracted Gabor features are encoded by the AR-LGC operator. The coded Gabor characteristic spectrum is divided into some blocks and each sub block is counted by histogram. These results are sequentially connected to form face expression feature vector, and using PCA algorithm to reduce dimension. At last SVM tools is adopted to classify and recognize the face. It is shown that the average recognition rate of the algorithm is 93.33%, which is 6.19% and 3.33% higher than that of AR-LGC+SVM and Gabor+PCA+SVM feature extraction algorithm respectively. This method can provide reference in face feature extraction.
Aiming at the incompleteness problem that the feature information extracted from Gabor wavelet and the dimension of disaster, an improved algorithm with face feature extraction based on two-dimensional Gabor wavelet and AR-LGC is proposed. In this algorithm, normalized face features are extracted by two-dimensional Gabor wavelet and the extracted Gabor features are encoded by the AR-LGC operator. The coded Gabor characteristic spectrum is divided into some blocks and each sub block is counted by histogram. These results are sequentially connected to form face expression feature vector, and using PCA algorithm to reduce dimension. At last SVM tools is adopted to classify and recognize the face. It is shown that the average recognition rate of the algorithm is 93.33%, which is 6.19% and 3.33% higher than that of AR-LGC+SVM and Gabor+PCA+SVM feature extraction algorithm respectively. This method can provide reference in face feature extraction.
引文
[1] 李健.基于Gabor变换和LBP相结合的人脸识别算法的研究[D].太原:太原理工大学,2016.LI Jian.Research of Face Recognition Based on Gabor Transform Combined with LBP Algorithm[D].Taiyuan:Taiyuan University of Technology,2016.
[2] 魏月纳.基于特征融合的人脸识别算法研究与应用[D].无锡:江南大学,2016.WEI Yuena.Research and Application of Face Recognition Algorithm Based on Feature Fusion [D].Wuxi:Jiangnan University,2016.
[3] LUO Yuan,YU Chaojing,ZHANG Yi,et al.Facial expression recognition based on fusion of extended LDP and Gabor features[J].The Journal of China Universities of Posts and Telecommunications,2018,25(1):48-53.
[4] 赖剑煌,阮邦志,冯国灿.频谱脸:一种基于小波变换和Fourier变换的人像识别新方法[J].中国图象图形学报,1999,4(10):811-817.LAI Jianhuang,YUEN P C,FENG Guocan.Spectroface:A wavelet-based and fourier-based approach for human face recognition[J].Journal of Image and Graphics,1999,4(10):811-817.
[5] ALOBAIDI T,MIKHAEL W B.Face recognition system based on features extracted from two domains[C]//2017 IEEE 60th International Midwest Symposium on Circuits and Systems(MWSCAS).[S.l.]:[s.n.],2017:977-980.
[6] LADES M,VORBRUGGEN J C,BUHMANN J M,et al.Distortion invariant object recognition in the dynamic link architecture[J].IEEE Transactions on Computers,1993,42(3):300-311.
[7] 嵇介曲.基于LBP算法的人脸识别研究[D].淮南:安徽理工大学,2017.JI Jiequ.Research on Face Recognition based on LBP Algorithm[D].Huainan:Anhui University of Science and Technology,2017.
[8] 刘东辉,卞建鹏,付平,等.支持向量机最优参数选择的研究[J].河北科技大学学报,2009,30(1):58-61.LIU Donghui,BIAN Jianpeng,FU Ping,et al.Study on the choice optimum parameters of support vector machine[J].Journal of Hebei University of Science and Technology,2009,30(1):58-61.
[9] 周慧敏,杨明.基于MBP算法和深度学习的人脸识别[J].河北工业科技,2019,36(1):25-30.ZHOU Huimin,YANG Ming.Face recognition based on monogenic binary patterns and deep learning[J].Hebei Journal of Industrial Science and Technology,2019,36(1):25-30.
[10] 许亚军,李玮欣.基于Gabor小波变换和神经网络的人脸识别研究[J].中国电子科学研究院学报,2017,12(5):534-539.XU Yajun,LI Weixin.Research on face recognition based on the Gabor wavelet and the neural network[J].Journal of China Academy of Electronics and Information Technology,2017,12(5):534-539.
[11] 胡正平,何薇,王蒙,等.Gabor调制的深度多层子空间人脸特征提取算法[J].信号处理,2017,33(3):338-345.HU Zhengping,HE Wei,WANG Meng,et al.Face feature extraction algorithm based on deep subspace with Gabor filter modulation[J].Journal of Signal Processing,2017,33(3):338-345.
[12] 程轶红.基于非对称局部梯度编码及多特征融合的人脸表情识别[D].合肥:合肥工业大学,2016.CHENG Yihong.Facial Expression Recognition Based on asymmetric Region Local Gradient Coding and Multi-feature Fusion [D].Hefei:Hefei University of Technology,2016.
[13] 胡敏,程轶红,王晓华,等.基于非对称局部梯度编码的人脸表情识别[J].中国图象图形学报,2015,20(10):1313-1321.HU Min,CHENG Yihong,WANG Xiaohua,et al.Facial expression recognition based on asymmetric local gradient coding[J].Journal of Image and Graphics,2015,20(10):1313-1321.
[14] 曹林,王东峰,刘小军,等.基于二维Gabor小波的人脸识别算法[J].电子与信息学报,2006,28(3):490-494.CAO Lin,WANG Dongfeng,LIU Xiaojun,et al.Face recognition based on two-dimensional Gabor wavelets[J].Journal of Electronics and Information Technology,2006,28(3):490-494.
[2] 魏月纳.基于特征融合的人脸识别算法研究与应用[D].无锡:江南大学,2016.WEI Yuena.Research and Application of Face Recognition Algorithm Based on Feature Fusion [D].Wuxi:Jiangnan University,2016.
[3] LUO Yuan,YU Chaojing,ZHANG Yi,et al.Facial expression recognition based on fusion of extended LDP and Gabor features[J].The Journal of China Universities of Posts and Telecommunications,2018,25(1):48-53.
[4] 赖剑煌,阮邦志,冯国灿.频谱脸:一种基于小波变换和Fourier变换的人像识别新方法[J].中国图象图形学报,1999,4(10):811-817.LAI Jianhuang,YUEN P C,FENG Guocan.Spectroface:A wavelet-based and fourier-based approach for human face recognition[J].Journal of Image and Graphics,1999,4(10):811-817.
[5] ALOBAIDI T,MIKHAEL W B.Face recognition system based on features extracted from two domains[C]//2017 IEEE 60th International Midwest Symposium on Circuits and Systems(MWSCAS).[S.l.]:[s.n.],2017:977-980.
[6] LADES M,VORBRUGGEN J C,BUHMANN J M,et al.Distortion invariant object recognition in the dynamic link architecture[J].IEEE Transactions on Computers,1993,42(3):300-311.
[7] 嵇介曲.基于LBP算法的人脸识别研究[D].淮南:安徽理工大学,2017.JI Jiequ.Research on Face Recognition based on LBP Algorithm[D].Huainan:Anhui University of Science and Technology,2017.
[8] 刘东辉,卞建鹏,付平,等.支持向量机最优参数选择的研究[J].河北科技大学学报,2009,30(1):58-61.LIU Donghui,BIAN Jianpeng,FU Ping,et al.Study on the choice optimum parameters of support vector machine[J].Journal of Hebei University of Science and Technology,2009,30(1):58-61.
[9] 周慧敏,杨明.基于MBP算法和深度学习的人脸识别[J].河北工业科技,2019,36(1):25-30.ZHOU Huimin,YANG Ming.Face recognition based on monogenic binary patterns and deep learning[J].Hebei Journal of Industrial Science and Technology,2019,36(1):25-30.
[10] 许亚军,李玮欣.基于Gabor小波变换和神经网络的人脸识别研究[J].中国电子科学研究院学报,2017,12(5):534-539.XU Yajun,LI Weixin.Research on face recognition based on the Gabor wavelet and the neural network[J].Journal of China Academy of Electronics and Information Technology,2017,12(5):534-539.
[11] 胡正平,何薇,王蒙,等.Gabor调制的深度多层子空间人脸特征提取算法[J].信号处理,2017,33(3):338-345.HU Zhengping,HE Wei,WANG Meng,et al.Face feature extraction algorithm based on deep subspace with Gabor filter modulation[J].Journal of Signal Processing,2017,33(3):338-345.
[12] 程轶红.基于非对称局部梯度编码及多特征融合的人脸表情识别[D].合肥:合肥工业大学,2016.CHENG Yihong.Facial Expression Recognition Based on asymmetric Region Local Gradient Coding and Multi-feature Fusion [D].Hefei:Hefei University of Technology,2016.
[13] 胡敏,程轶红,王晓华,等.基于非对称局部梯度编码的人脸表情识别[J].中国图象图形学报,2015,20(10):1313-1321.HU Min,CHENG Yihong,WANG Xiaohua,et al.Facial expression recognition based on asymmetric local gradient coding[J].Journal of Image and Graphics,2015,20(10):1313-1321.
[14] 曹林,王东峰,刘小军,等.基于二维Gabor小波的人脸识别算法[J].电子与信息学报,2006,28(3):490-494.CAO Lin,WANG Dongfeng,LIU Xiaojun,et al.Face recognition based on two-dimensional Gabor wavelets[J].Journal of Electronics and Information Technology,2006,28(3):490-494.