摘要
针对目前光纤绕制缺陷检测方法准确率低,实用性差等问题,提出一种基于低秩表示技术的光纤环绕制缺陷检测方法。首先,利用光纤环图像纹理的重复性和缺陷的局部性特征,将无缺陷的光纤环图像建模为低秩结构,将缺陷建模为稀疏结构。然后,基于低秩表示理论对绕制缺陷检测问题进行模型构建,并利用增广拉格朗日乘子法求解其目标函数,以近似恢复原始图像的行空间并分离出缺陷区域。通过实验对算法进行验证,结果表明,提出的光纤绕制缺陷检测方法是可行的,能够有效检测出缺陷,且与其他算法相比,本算法取得缺陷检测精度最高。
Aiming at the problems of low accuracy and poor practicability of current optical fiber winding defect detection methods, a method of optical fiber winding defect detection based on low-rank representation technology is proposed. Firstly, for the repetitiveness of the fiber coil image texture and the local characteristics of defects, the algorithm models defect-free fiber coil image as a low-rank structure and the defects as a sparse structure. Then, based on low-rank representation, the problem of winding defect detection is modeled, and the Augmented Lagrangian multiplier method(ALM) is used to approximately restore line space of original image and isolate defect area. The algorithm is verified through experiments, and the results show that the fiber winding defect detection method proposed in this paper is feasible and can effectively detect defects. Compared with other algorithms, the algorithm in this paper achieves the highest defect detection accuracy.
作者
陈晓乐
杨瑞峰
郭晨霞
CHEN Xiaole;YANG Ruifeng;GUO Chenxia(School of Instrument and Electronics,North University of China,Taiyuan 030051,China;Automation Testing Equipment and System Engineering Technology Research Center,Taiyuan 030051,China)
出处
《激光杂志》
CAS
北大核心
2022年第2期48-52,共5页
Laser Journal
基金
山西省归国留学人员科研资助项目(No.2020-111)。
关键词
光纤环绕制
缺陷检测
低秩表示
奇异值分解
fiber coil winding
low rank representation
defect detection
singular value decomposition
