摘要
利用混沌对初值的极端敏感依赖性,可以对仅有微小差别的模式进行识别。提出一种基于模糊混沌神经网络的算法,并应用到人脸识别中。由于引入了混沌噪声,可使网络具有很强的抗干扰能力,能有效避免人脸图像光照、姿态等因素对人脸识别的影响,也避免了复杂的特征提取工作。利用ORL人脸图像数据库进行了仿真实验,结果表明,混沌神经网络算法精度高、迭代步骤少、收敛快,混沌神经网络应用于人脸识别是有效的,能提高识别率。
For its sensitive dependence with the Initial value, chaos can be applied to the pattern recognition of the ones with extremely small difference. An algorithm based on chaotic neural network was proposed and used for face recognition. For introducing chaotic noise, the network obtains a better anti-jamming. It can avoid being affected by the factors such as illumination and gesture. And many complex feature extractions can be avoided. Experimental results based on 0RL face database show that the precision of the chaotic neural network algorithm is higher and the iteration steps are fewer and the speed of convergence is quicker. Chaotic neural network used for face recognition is effective and it can enhance recognition rate.
出处
《计算机应用》
CSCD
北大核心
2008年第6期1549-1551,1558,共4页
journal of Computer Applications
关键词
神经网络
混沌
模糊
人脸识别
模式识别
neural network
chaotic
fuzzy
face recognition
pattern recognition