用填充函数法改进的人脸比对算法  

Face Comparison Algorithm Based on Filled Function Method

作　　者：钟以维[1] 徐应涛[2] 张莹[2] 

机构地区：[1]浙江广播电视大学三门学院,浙江台州317100 [2]浙江师范大学数理与信息工程学院,浙江金华321004

出　　处：《计算机技术与发展》2009年第8期78-81,共4页Computer Technology and Development

基　　金：国家自然科学基金项目(10571116)

摘　　要：人脸识别技术在国家安全、金融、执法、司法等领域都有广泛应用,其中比对分类算法是人脸识别系统的重要算法。目前所用BP神经网络的分类方法收敛速度慢、易陷入局部最优,而填充函数法是确定型全局优化方法,能够快速跳出当前局部极小点,得到一个更低的极小点,重复此过程得到全局极小点。用填充函数法改进的BP神经网络实现分类器的方法可以提高算法的收敛速度,降低误判率,增强其全局寻优的能力。实验表明该人脸比对算法能够提高比对准确率,减少计算量,提高比对速度。Face recognition serves national security, finance, law enforcement, etc. ,and face comparison algorithm is an important component in face recognition system. Aiming at the handicaps in current methods such as slow convergence or easiness of getting into local optimization,works out an improved BP neural network classifier combining the filled function method. The filled function method can find a lower local minimizer by leaving the local minimizer previously found, and by repeating these processes, a global minimizer can be obtained. It can enhance the convergence speed, reduce the error recognition rate, and improve the ability of finding the global solution. Experiments show that this face comparison algorithm enhances the accuracy of comparison precision, decreases the amount of calculation and improves the speed of comparison.

关 键 词：人脸比对 填充函数 分类器 BP神经网络 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] O221[自动化与计算机技术—计算机科学与技术]