基于Adaboost改进算法的铸坯表面缺陷检测方法

钢铁研究学报, 2012, 24(9): 59-62.

1.上海交通大学图像处理与模式识别研究所,上海200030

2.上海交通大学图像处理与模式识别研究所,上海200030

3.宝山钢铁股份有限公司研究院,上海201900

4.宝山钢铁股份有限公司研究院,上海201900

基金项目: 国家自然科学基金资助项目(60873137)

Surface Defect Detection of Slab Based on the Improved Adaboost Algorithm

针对钢铁铸坯表面检测及缺陷识别问题,从图像处理及机器学习角度,提出一种基于Adaboost算法的进行钢铁铸坯表面缺陷检测,并结合Gabor小波和Canny边缘检测进行处理,排除伪缺陷的新方法.大量试验表明:该方法能够较好地检出具有缺陷的钢铁铸坯,且具有准确率高、速度快、易实施等优点.

[1]	Freund Y;Schapire R E.A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting[A].Barcelona:[s.n.],1995:23.
[2]	Viola P;Jones M.Rapid Object Detection Using a Boosted Cascade of Simple Features[A].Kauai,USA:IEEE Press,2001:511.
[3]	WANG Y;SUN Y;L(U) P.Detection of Line Weld Defects Based on Multiple Thresholds and Support Vector Machine[J].NDT & E International,2005(05):4.
[4]	H.I. Shafeek;E.S. Gadelmawla;A.A. Abdel-Shafy .Automatic inspection of gas pipeline welding defects using an expert vision system[J].NDT & E international: Independent nondestructive testing and evaluation,2004(4):301-307.
[5]	刘清坤,阙沛文,宋寿鹏.基于支持向量机的石油管线缺陷识别方法研究[J].传感器技术,2005(03):30-31,34.
[6]	郭捷,施鹏飞.基于颜色和纹理分析的车牌定位方法[J].中国图象图形学报A辑,2002(05):472-476.
[7]	Jacobsen C;Zscherpel U;Perner P.A Comparison Between Neural Networks and Decision Trees[A].New York:springer-verlag,1999:144.

上一张下一张

计量

作者相关

文章评分