【摘要】 针对硅钢酸洗-冷连轧生产过程中钢带表面缺陷的特点,在提取钢带图像低层属性特征的基础上,采用决策树、SVM、BP神经网络等不同的模式识别方法对钢带表面缺陷进行分类研究,试验过程包含训练集与测试集的收集、特征提取、机器学习与分类、结果分析。该研究以提高硅钢冷轧生产过程的自动化水平为目标,依据试验结果提出缺陷识别的优化方法,并应用于武钢硅钢冷轧生产线,提高了自动化水平。更多还原

【Abstract】 According to the feature of surface defect image in silicon strip Pickling-CC Rolling process,the different classification method study has been launched on the using of Decision Tree,SVM,BPNeural Network and on the Extraction of image lower-attributes. Experiment include that collection of training set and test image,extraction of the image feature,Machine Learning and Classification,analysis of results. The target of this research is raising the level of silicon strip’s automatic production. At last we have proposed optimization method of Defect Detection,and application it on WISCO for the level of automatic production.更多还原