摘要
介绍了4种计算智能方法,即神经网络、模糊计算、遗传算法和免疫算法,通过总结和比较各智能算法的优缺点及其在各个领域的应用,给出了计算智能存在的主要问题及发展趋势。设计了脱机手写签名识别软件,提取签名图像的骨架特征和边角特征,并利用BP神经网络方法有效地识别真、假签名。
This paper introduces four computing intelligence methods,namely,neural network,fuzzy computation,genetic algorithm and immune algorithm,compares the advantages and disadvantages of intelligent algorithm and its application in all fields,and gives the main problems and development trends in computational intelligence.A off-line handwritten signature recognition software is designed,skeleton features,edge and angle features are extracted,and signature recognition is effectively realized by BP neural network.
出处
《电脑开发与应用》
2012年第10期1-3,共3页
Computer Development & Applications
关键词
计算智能
神经网络
模糊计算
遗传算法
免疫算法
computing intelligence
neural network
fuzzy computation
genetic algorithm
immune algorithm
