摘要
针对人脸识别中,利用粒子群算法训练支持向量机进行分类识别时存在易陷入局部最优和收敛速度慢的问题,提出一种基于雁群优化算法的人脸识别方法。将主成分分析与独立成分分析相结合提取人脸特征,利用支持向量机进行分类,在分类识别的过程中,引入雁群优化算法以提高速度和效率。实验结果表明,与标准粒子群算法相比,改进的粒子群算法提高了人脸识别率,具有较快的识别速度。
Because the particle swarm optimization may easily fall into local optimum and takes long runtime while training support vector machine as a classifier for human face recognition,ageese swarm optimization algorithm was proposed.The principle component analysis and the independent component analysis were combined to extract human face features,and the support vector machine was used as a classifier.To achieve higher speed and recognition accuracy,the geese swarm optimization algorithm was introduced in the classification stage.The experimental results show that the human face recognition system using the improved particle swarm optimization algorithm was improved in not only the recognition rate but also the efficiency compared with the standard particle swarm optimization.
出处
《计算机工程与设计》
CSCD
北大核心
2014年第12期4302-4305,4310,共5页
Computer Engineering and Design
基金
山西省自然科学基金项目(2013011016-1)
教育部博士点基金项目(2011081047)
关键词
人脸识别
主成分分析
独立成分分析
支持向量机
粒子群算法
face recognition
principal component analysis（PCA）
independent component analysis（ICA）
support vector machine（SVM）
particle swarm optimization（PSO）
