摘要
运用神经网络处理非线性问题的优势,将其应用于带钢表面缺陷的识别与分类研究。本文采用灰度共生矩阵的特征提取,提出了基于BP神经网络进行缺陷识别与分类的方法,编制了带钢表面缺陷的识别与分类软件。分类测试表明,该软件有较好的识别与分类效果。
Taking the advantage of neural network in dealing with non-linear problem, neural network is applied to study the discernment and classification of the strip surface defects. A discernment and classification method is proposed based on BP neural network,which makes use of grey level intergrowth matrix. The software for discernment and classification of banding strip surface defects was written. Classification test result shows that this software has good discernment and classification performance.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2006年第12期1692-1694,共3页
Chinese Journal of Scientific Instrument
基金
国家科技部重大基础研究前期研究专项资金(2003CCA03900)
国家自然科学基金(50574019)
上海宝钢集团公司联合资助项目
关键词
带钢表面缺陷
识别与分类
灰度共生矩阵
人工神经网络
BP算法
banding strip surface defect discern and classification grey level intergrowth matrix artificial neural network BP algorithm
