基于多特征融合和Bagging神经网络的人耳识别

提出了一种多特征信息融合的人耳识别方法。应用Zernike矩方法和非负矩阵分解法(NMF)分别提取出具有旋转不变性的人耳几何特征和人耳子空间投影系数特征，将这两种具有一定互补性的特征串行融合，得到一个分类能力更强的特征。在此基础上，采用神经网络进行人耳识别，为了提高了神经网络的分类准确率和泛化能力，采用Bagging方法构造了Bagging神经网络。给出了一些对比实验，结果验证了方法具有较高识别率。

Ear recognition based on feature fusion and Bagging neural network

A new ear recognition method based on feature fusion was presented. Firstly, ear projection coefficient features were extracted by NMF methods. Then these features were combined with the ear rotation invariance Zemike moments features so as to get a new feature which had higher discriminating power. With this new feature, a Bagging neural network was built up by using bagging algorithm that improved classification accuracy and generalization of neural network. At last, the results of some comparative experiments show that this ear recognition method can get higher recognitiori rate.