摘要
该文使用神经网络技术对人脸的特征提取及识别做了研究。在降低图像解析度的基础上使用神经网络来进行特征压缩,可以有效地降低特征维数。使用神经网络分类器对压缩后的特征进行分类,分类结果以隶属度来表征待识别人脸属于各类别的可能性。实验结果证明了该方法的有效性。
In this paper,neur al network technology is used for face feature extraction and recognition.Based o n resolution reduction,neural network is then utilized for feature compression ,which can effectively reduce the dimension of features.Then the compressed fe atures are classified by neural network classifier,which can classify a given p attern to more than one class simultaneously with a membership degree denotes it s possibility.Simulation results demonstrate the effectiveness of this approach .
出处
《计算机工程与应用》
CSCD
北大核心
2002年第11期33-34,37,共3页
Computer Engineering and Applications
基金
国家自然科学基金资助(编号:60072034)
