基于粗糙-模糊神经网络的焊接图像缺陷识别  

Defect Recognition of Welding Image Based on Rough-Fuzzy Network

作　　者：高顶[1] 张长明[1] 李国庆[1] 张晓光[1] 

机构地区：[1]中国矿业大学机电工程学院,徐州221008

出　　处：《华东理工大学学报（自然科学版）》2006年第9期1126-1129,共4页Journal of East China University of Science and Technology

基　　金：江苏省博士后科研基金资助项目(0401016B);中国矿业大学科技基金资助项目(2005B005)

摘　　要：针对焊接图像缺陷识别中提取的特征受噪声干扰比较严重以及现有的识别算法准确率低的问题,提出了一种基于粗糙模糊神经网络的缺陷识别算法。该算法充分利用了粗糙集的属性约简、模糊集的处理不精确数据以及神经网络的自学习、对任意函数逼近的优点,有效地解决了不确定建模过程中样本数据受到噪声干扰、模型结构难以确定的问题。仿真结果表明:该算法能有效地提高焊缝图像的缺陷识别能力。To solve the problem that defect characteristics extracted from welding images are seriously disturbed by noises and lower precision of recognition algorithm, a defect recognition algorithm based on rough fuzzy neuron networks is developed. The algorithm takes full advantages of reduction of attributes of rough sets, uncertain data processing of fuzzy sets, self-learning of neuron networks, and approximation to any function. It can solve the problems in the course of uncertainly modeling, such as noise disturbance ,difficulty in determining model structure. The simulation result shows that the algorithm can effectively improve the recognition of welding images.

关 键 词：粗糙集 模糊集 神经网络 焊接缺陷 识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]