基于Haar-Lbp特征和粒子群神经网络的人耳识别

Ear Recognition Based on Haar-lbp Ear Feature and Pso-Bp Neural network

在分析人耳Haar和Lbp特征的基础上,提出了一种基于神经网络和人耳Haar-Lbp特征对人耳进行识别的方法.文中选取了人耳图像的Haar和Lbp双特征值作为神经网络的训练样本,并尝试利用粒子群优化算法与BP结合算法训练网络.仿真结果表明,文中提出的方法取得较好的效果.

On the basis of analysis of the the Haar characteristics and the Lbp characteristics of the human ear, the proposed method based on neural network and human ear Haar - Lbp characteristics of the human ear recognition. Select the texture characteristics of the human ear image value as a neural network training samples, and try to take advantage of the particle swarm optimization algorithm combined with BP algorithm to train the network. The simulation results show that the proposed method achieved very good results.