摘要
在圆柱形覆膜锂电池质量检测中,圆周面破膜缺陷影响着产品安全性,故研究一种高速度、高精度的圆周面破膜缺陷检测方法十分必要。针对这一安全问题,笔者提出了一种利用中值滤波图像与原始图像相减获取差值图像的方法。由于光照原因,破膜缺陷处薄膜表面的变化在差值图像中会呈现出明暗变化,根据这种明暗变化,采用阈值分割的方法对缺陷进行提取与分类。实验表明,该方法能够有效检测破膜缺陷,平衡光照和不同包膜颜色引起的灰度值变化,避免圆周表面黑色污点的误检。算法适用性强,检测效率高。
In the quality inspection of cylindrical film-covered lithium batteries,the defect of the peripheral surface film affects the safety of the product,so it is necessary to study a high-speed and high-precision detection method of the peripheral surface film defect.Aiming at this security problem,a method for subtracting the median filtered image from the original image to obtain a difference image is proposed.Due to the reason of illumination,the change of the film surface at the film defect will show a light and dark change in the difference image.Based on this light and dark change,threshold segmentation is used to extract and classify defects.Experiments show that the method can effectively detect film-breaking defects,balance the gray value changes caused by illumination and different envelope colors,avoid the false detection of black stains on the circumferential surface,and the algorithm has strong applicability and high detection efficiency.
作者
苑玮琦
李岩
Yuan Weiqi;Li Yan(Computer Vision Group,Shenyang University of Technology,Shenyang Liaoning 110870,China;Key Laboratory of Machine Vision,Shenyang Liaoning 110870,China)
出处
《信息与电脑》
2020年第2期108-110,共3页
Information & Computer
关键词
锂电池
圆周面破膜
中值滤波
图像相减
阈值分割
lithium battery
peripheral surface breaking film
median filtering
image subtraction
threshold segmentation
