基于改进遗传算法和BP神经网络的人脸识别方法
|
摘要
人脸识别因其高可靠性在信息安全领域得到了广泛的应用与研究。为了提高识别的准确度,提出了一种基于改进遗传算法和BP神经网络的人脸识别方法。利用主成分分析法对人脸图像特征向量进行提取压缩,再利用线性判别分析提取人脸类别信息,作为BP神经网络的输入。对遗传算法选择算子进行改进,在排序选择策略基础上引入最优保存策略,同时采用自适应交叉、变异方式,得到一种改进的遗传算法。采用该改进遗传算法训练BP神经网络,实现人脸识别。基于ORL人脸数据库的仿真结果表明,该算法实现简单、准确率高,对人脸识别算法的研究具有重要意义。
The high reliability of face recognition makes it widely applied and researched in the field of information security.In order to improve the accuracy of face recognition,a face recognition method based on improved genetic algorithm and BP neural network is proposed. The method uses principal component analysis to extract the compressed image features vectors,and then uses linear discriminant analysis to extract the face category information,as the input to the neural network. To improve the selection operator of genetic algorithm,optimal selection strategy is combined with the sort selection strategy,also the adaptive crossover and mutation probability are introduced to obtain an improved genetic algorithm,the improved genetic algorithm is used to train the BP neural network to realize face recognition. The experimental results on ORL face database show that the method introduced in this paper is simple and accurate,and it is important to study the face recognition algorithm.
The high reliability of face recognition makes it widely applied and researched in the field of information security.In order to improve the accuracy of face recognition,a face recognition method based on improved genetic algorithm and BP neural network is proposed. The method uses principal component analysis to extract the compressed image features vectors,and then uses linear discriminant analysis to extract the face category information,as the input to the neural network. To improve the selection operator of genetic algorithm,optimal selection strategy is combined with the sort selection strategy,also the adaptive crossover and mutation probability are introduced to obtain an improved genetic algorithm,the improved genetic algorithm is used to train the BP neural network to realize face recognition. The experimental results on ORL face database show that the method introduced in this paper is simple and accurate,and it is important to study the face recognition algorithm.
引文
[1]樊春玲,陈秀霆.基于PCA和LDA的人脸识别系统设计[J].控制工程,2012,19(4):165-168.
[2]伊力哈木·亚尔买买提.基于改进型PCA和LDA融合算法的人脸图像识别[J].计算机仿真,2013,30(1):415-418.
[3]沈凌云,郎百和,朱明.一种基于人工神经网络的人脸识别方法[J].液晶与显示,2011,26(6):836-840.
[4]杨洪军,董玉华.基于GA-BP算法的超声波测量精度优化研究[J].计算机工程与科学,2016,38(5):1066-1070.
[5]李海朋,李晶皎,闫爱云,等.人脸识别中的遗传神经网络并行实现[J].计算机科学,2015,42(6A):168-174.
[6] GAO Y F. Genetic algorithm for optimizing the layout of logistics park based on shortest path and elitist strategy[J]. Journal of Information and Computational Science,2015,12(10):3765-3774.
[7]桂预风,程战员.基于f-k优化近邻法的PCA人脸识别算法[J].武汉理工大学学报(信息与管理工程版),2013,35(6):820-824.
[8]万源,周达丽,童恒庆. LLE与核Fisher判别分析结合的人脸识别研究[J].武汉理工大学学报(信息与管理工程版),2013,35(6):799-783.
[9]谈书才,黄景涛.基于LLE和LS-SVM的人脸识别方法[J].微电子学与计算机,2010,27(7):110-113.
[10]褚勤.基于小波分析和支持向量机的人脸识别方法研究[D].广州:华南理工大学,2012.
[11]张锦,成奋华,林雪梅,等.基于子图特征组合的人脸识别技术研究[J].湖南大学学报(自然科学版),2007,34(6):70-73.
[2]伊力哈木·亚尔买买提.基于改进型PCA和LDA融合算法的人脸图像识别[J].计算机仿真,2013,30(1):415-418.
[3]沈凌云,郎百和,朱明.一种基于人工神经网络的人脸识别方法[J].液晶与显示,2011,26(6):836-840.
[4]杨洪军,董玉华.基于GA-BP算法的超声波测量精度优化研究[J].计算机工程与科学,2016,38(5):1066-1070.
[5]李海朋,李晶皎,闫爱云,等.人脸识别中的遗传神经网络并行实现[J].计算机科学,2015,42(6A):168-174.
[6] GAO Y F. Genetic algorithm for optimizing the layout of logistics park based on shortest path and elitist strategy[J]. Journal of Information and Computational Science,2015,12(10):3765-3774.
[7]桂预风,程战员.基于f-k优化近邻法的PCA人脸识别算法[J].武汉理工大学学报(信息与管理工程版),2013,35(6):820-824.
[8]万源,周达丽,童恒庆. LLE与核Fisher判别分析结合的人脸识别研究[J].武汉理工大学学报(信息与管理工程版),2013,35(6):799-783.
[9]谈书才,黄景涛.基于LLE和LS-SVM的人脸识别方法[J].微电子学与计算机,2010,27(7):110-113.
[10]褚勤.基于小波分析和支持向量机的人脸识别方法研究[D].广州:华南理工大学,2012.
[11]张锦,成奋华,林雪梅,等.基于子图特征组合的人脸识别技术研究[J].湖南大学学报(自然科学版),2007,34(6):70-73.