Strip Steel Surface Defect Recognition Based on Complex Network Characteristics
-
摘要: 针对带钢表面缺陷识别问题,提出一种基于动态演化复杂网络特性的特征描述方法, 这些特征同时具有位移、旋转不变性、大小不变性、较强的抗干扰能力和鲁棒性,为 缺陷识别提供良好的分类特征;为了提高分类器的效率,应用主成分分析法 (Principal component analysis, PCA) 对复杂网络特 征向量进行特征降维处理;采用最优有向无环图支持向量机 (Directed acyclic graph support vector machine, DAG-SVM)算法进行缺陷分类.结果表明该方法识别率高而且识别速度快.
-
关键词:
- 缺陷识别 /
- 复杂网络特征 /
- 主成分分析法 /
- 有向无环图支持向量机
Abstract: A feature extraction method based on the characteristics of dynamic evolution complex networks is proposed for the strip steel surface defect recognition. The extracted features possess displacement, rotation and size invariability, strong anti-interference ability and robustness, therefore they are good classification features for steel surface defect recognition. In order to improve the efficiency of classification, the principal component analysis (PCA) is adopted to reduce the dimension of the feature vector. The directed acyclic graph support vector machine (DAG-SVM) algorithm is used for the defect classification. The experimental results show that this method is of high recognition rate and fast recognition speed. -
[1] Choi K, Koo K, Lee J S. Development of defect classification algorithm for POSCO rolling strip surface inspection system. In: Proceedings of the International Joint Conference on SICE-ICASE. Busan, Korea: IEEE, 2006. 2499-2502[2] Peng K X, Zhang X L. Classification technology for automatic surface defects detection of steel strip based on improved BP algorithm. In: Proceedings of 5th International Conference on Natural Computation. Tianjin, China: IEEE, 2009. 110-114[3] Han Ying-Li, Yan Yun-Hui. Discernment and classification of banding strip surface defect based on BP neural network. Chinese Journal of Scientific Instrument, 2006, 27(12): 1692-1694(韩英莉, 颜云辉. 基于BP神经网络的带钢表面缺陷的识别与分类. 仪器仪表学报, 2006, 27(12): 1692-1694)[4] Zhang Yuan, Chen Wan-Sheng, Zhao Jie. Classification of surface defects of strips based on invariable moment functions. Opto-Electronic Engineering, 2008, 35(7): 90-94(张媛, 程万胜, 赵杰. 不变矩法分类识别带钢表面的缺陷. 光电工程, 2008, 35(7): 90-94)[5] Barabasi A L. Scale-free networks: a decade and beyond. Science, 2009, 325(5939): 412-413[6] Huang Wen-Liang, Liu Yong, Zhong Zhi-Qiang, Shen Zhong-Ming. Complex network based SMS filtering algorithm. Acta Automatica Sinica, 2009, 35(7): 990-996(黄文良, 刘勇, 钟志强, 沈仲明. 基于复杂网络的垃圾短信过滤算法. 自动化学报, 2009, 35(7): 990-996)[7] He Dong-Xiao, Zhou Xu, Wang Zuo, Zhou Chun-Guang, Wang Zhe, Jin Di. Community mining in complex networks-clustering combination based genetic algorithm. Acta Automatica Sinica, 2010, 36(8): 1160-1170(何东晓, 周栩, 王佐, 周春光, 王喆, 金弟. 复杂网络社区挖掘——基于聚类融合的遗传算法. 自动化学报, 2010, 36(8): 1160-1170)[8] Backes A R, Casanova D, Bruno O M. A complex network-based approach for boundary shape analysis. Pattern Recognition, 2009, 42(1): 54-67[9] Verein Deutscher Eisenhuttenleute [Author], Chinese Society for Metals [Translator]. Hot Rolled, Cold Rolled, Hot Plating and Electroplating Flat Steel Product Surface Defect Picture Spectrum. Beijing: Iron and Steel Editor Office, 2000. 6-37(德国钢铁学会 [著], 中国金属学会 [译]. 热轧、冷轧、热镀、电镀金属板带的表面缺陷图谱. 北京: 中国金属学会《钢铁》编辑部, 2000. 6-37)[10] Wallace T P, Wintz P A. An efficient three-dimensional aircraft recognition algorithm using normalized Fourier descriptors. Computer Graphics and Image Processing, 1980, 13(2): 99-126[11] Chuang G C H, Kuo C C J. Wavelet descriptor of planar curves: theory and applications. IEEE Transactions on Image Processing, 1996, 5(1): 56-70[12] Martinez A M, Kak A C. PCA versus LDA. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2001, 23(2): 228-233[13] Li Jun-Tao, Jia Ying-Min. Huberized multiclass support vector machine for microarray classification. Acta Automatica Sinica, 2010, 36(3): 399-405[14] Qian Kun, Ma Xu-Dong, Dai Xian-Zhong, Hu Chun-Hua. Optimal DAGSVM based posture recognition for human-robot interaction. Journal of Image and Graphics, 2009, 14(1): 118-124(钱堃, 马旭东, 戴先中, 胡春华. 基于最优DAGSVM的服务机器人交互手势识别. 中国图象图形学报, 2009, 14(1): 118-124)
点击查看大图
计量
- 文章访问数: 2498
- HTML全文浏览量: 74
- PDF下载量: 978
- 被引次数: 0