摘要
针对锂电池正负极冗余度缺陷的检测要求,应用机器视觉技术,研究一种在线、高效的检测算法。从X射线原始图像中提取需要检测的感兴趣区域(ROI),对ROI图像进行频域降噪处理,然后采用空域滤波、形态学操作、轮廓提取等一系列空域操作,将负极片冗余区域提取出来,再运用直线检测技术找出线段信息,排除干扰项,找到目标冗余线段,提取其长度和角度等特征信息,设计分类器,将正负极冗余度等不满足技术要求的锂电池识别出来。实验表明,该算法准确率达98.5%、平均检测时间为721 ms,具有稳定性好、运行速度快等优点,满足工业检测中的实际需要。
Aiming at the detection requirements for redundancy defects of positive and negative electrodes,an on-line and efficient detection algorithm is studied by using machine vision technology.Extract the region of interest(ROI)from the original X-ray image,then denoise the ROI image in frequency domain.Extract the redundant region of lithium battery anode after a serial of operations in space domain,such as spatial filtering,morphological operation,contour extraction and so on.Find the targeted redundant segments without interference by using line detection technology.Extract the characteristic information and design a classifier.The results of the experiment show that the accuracy of algorithm is 98.5%,as well as the average detection time is 721 ms.The algorithm satisfies the actual needs of industrial detection with its good stability and fast running speed.
作者
陈家智
吴永明
CHEN Jia-zhi;WU Yong-ming(School of Electromechanical Engineering,Guangdong University of Technology,Guangzhou 510006,China)
出处
《组合机床与自动化加工技术》
北大核心
2021年第12期75-78,共4页
Modular Machine Tool & Automatic Manufacturing Technique
