基于改进模拟退火算法的手写体数字识别  

Handwritten Digit Recognition Based on Improved Simulated Annealing Algorithm

作　　者：洪沛霖[1] 张佑生[1] 邢燕[1] 

机构地区：[1]合肥工业大学计算机与信息学院,安徽合肥230009

出　　处：《计算机技术与发展》2007年第9期15-17,20,共4页Computer Technology and Development

基　　金：教育部高等学校博士学科点专项科研基金(20050359012)

摘　　要：对模糊C-均值聚类作了简要分析和评论,在此基础上将模拟退火机制引入其中,以克服模糊C-均值聚类的局部性和对初始聚类中心的敏感性;采用了基于贴近度和择近原则的模糊识别方法;详细设计了算法。仿真结果说明,该方法在识别速度和准确率方面都达到了令人满意的效果,为手写体数字的识别提供了一种新方法,也拓展了模糊理论的应用范围。Fuzzy C - means clustering is introduced and remarked. On the basis of systematic analysis of current algorithms,simulated annealing rnechanism is inducted into fuzzy clustering to solve the locality and the sensitiveness of the initial condition of fuzzy C - means clustering. Then, proposes the fuzzy discern method based on approximation value and the principle of selecting the near. Finally, the algorithm is designed in detail. Simulation results show that the new method gets a satisfied keywords result both application on speed and the correct rate. Thus a new handwritten digit recognition method is presented, and it widens the application of fuzzy theory.

关 键 词：模拟退火算法 模糊模式 数字识别 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]