计算机技术与发展Issue(9)：70-73,4.DOI:10.3969/j.issn.1673-629X.2013.09.018
基于Matlab和C#的数显仪表数字识别系统
Digital Recognition System of Numerical Instruments Based on Matlab and C#
摘要
Abstract
Digital recognition technology is widely applied in numerical instruments. Improving its recognition rate is helpful for increasing the automation level of instruments. In this paper,Otsu algorithm is used to realize the image binarization,isolated pixel connected domain is adopted to eliminate noise,vertical projection algorithm is applied to accomplish the character segmentation and modular algorithm is exploited to extract the feature. Furthermore,3-layered BP neural network is established,and the adaptive learning factor with a momen-tum term is used to adjust the parameters of BP neural network. Based on dynamic link library method,the composite Matlab and C# dig-ital recognition system is designed. The results indicate that the recognition rate can reach 98% for single-number images and 92. 5% for multi-number images.关键词
数字识别/神经网络/特征提取/数显仪表Key words
digital recognition/neural network/feature extraction/numeral instruments分类
信息技术与安全科学引用本文复制引用
张宁,秦德鑫,王秀芳..基于Matlab和C#的数显仪表数字识别系统[J].计算机技术与发展,2013,(9):70-73,4.基金项目
“十二五”国家科技支撑重点项目(2012BAH12B00) （2012BAH12B00）
黑龙江省教育科学技术研究项目(12521056) （12521056）
