基于小波变换和群体决策方法识别人脸图像

FACE RECOGNITION BASED ON WAVELET TRANSFORM AND GROUP DECISION MAKING APPROACH

文中使用Ｄａｕｂｅｃｈｉｅｓ正交小波变换对人脸图像作预处理，得到它在不同频带上的４个子图像，对它们分别提取奇异值特征，然后用最近邻方法进行分类，这样使得４组分类结果之间相关性减少，差异性增大．同时设计了一种适用于多分类结果融合的群体决策算法，并且对分类结果有选择地进行融合．实验表明，对于奇异值特征而言，作小波变换预处理得到的４个子图像的分类结果之间具有一定互补性，用群体决策算法作融合，可以提高其分类性能．实验也说明，群体决策方法计算复杂度低，优于常用的计分法的融合效果．

Daubechies orthogonal wavelet transform is used to preprocess the face image,resulting in its four subimages belonging to different frequency bands.The singular value feature vectors are then extracted from the four subimages respectively,and the nearest neighbour classifier is used to recognize them.This makes correlation reduce and makes difference raise between the four groups of sort results.A group decision making algorithm is designed to combine the multiple sort results.Moreover,all of or part of them can be combined according to group consensus index.Experiments show that so far as singular value feature is concerned,there is a certain complementarity between the sort results of the four subimages.The sort performance can be improved using the group decision making algorithm,which has lower computational complexity and better combination results than the commonly used mark counting approach.