基于Hopfield网络的手写体汉字识别系统  

Handwritten Chinese Character Recognition System Based on Hopfield Network

作　　者：崔泽乾 白云杰 韩阳 陈丽芳 CUI Ze-qian;BAI Yun-jie;HAN Yang;CHEN Li-fang(College of Metallurgy and Energy,North China University of Science and Technology,Tangshan Hebei 063210,China;College of Science,North China University of Science and Technology,Tangshan Hebei 063210,China;Department of Discipline Construction,North China University of Science and Technology,Tangshan Hebei 063210,China;The Key Laboratory of Engineering Computing in Tangshan,College of Science of North China University of Science and Technology,Tangshan Hebei 063210,China)

机构地区：[1]华北理工大学冶金与能源学院,河北唐山063210 [2]华北理工大学理学院,河北唐山063210 [3]华北理工大学学科建设处,河北唐山063210 [4]唐山市工程计算重点实验室,华北理工大学理学院,河北唐山063210

出　　处：《华北理工大学学报（自然科学版）》2021年第2期123-131,共9页Journal of North China University of Science and Technology：Natural Science Edition

基　　金：河北省重点研发计划项目(NO.20270902D)。

摘　　要：针对基于Hopfield网络算法实现的手写汉字体识别系统,使用Python语言中Numpy和Cv2库对数据进行预处理,通过Neurolab库训练Hopfield神经网络,编写相似度对比算法对其进行改进,采用Pyqt5实现功能的可视化。研究结果表明,该汉字体识别系统具有图片识别和手写识别两大功能。采用改进后的Hopfield神经网络算法,该系统的识别效率有了很大的提升,识别准确度保持在80%以上,并且具有良好的兼容性。Aiming at Chinese handwriting recognition based on Hopfield network algorithm in the market.Numpy and Cv2 libraries in Python language were used to preprocess the data,the Neurolab library was used to train Hopfield neural network,the similarity comparison algorithm was written to improve it,and Pyqt5 was used to realize the visualization of functions.The results show that Chinese font recognition system has two functions such as picture recognition and handwriting recognition.By using the improved Hopfield neural network algorithm,the recognition efficiency of the system has been greatly improved,and its recognition accuracy remains above 80%with good compatibility.

关 键 词：汉字识别 HOPFIELD神经网络 深度学习 图像分割 

分 类 号：TP391.43[自动化与计算机技术—计算机应用技术]