摘要
提出一种针对无交叉的纤维自动测量长度的方法。利用图像的边缘检测技术分离背景与纤维,设置阈值剔除杂质,通过对纤维细化得到其单像素宽度的骨架,追踪骨架的像素链,统计像素个数,最后通过标定的长度和实际比例计算得到纤维长度。试验结果表明,针对背景不复杂的无交叉纤维图像,该自动测量纤维长度的方法具有较好的鲁棒性。
A method for automatically measuring the length of fibers without crossover is proposed.Image edge detection technology is used to separate the background and fibers,a threshold is set to remove impurities,the single-pixel width skeleton is obtained by thinning the fibers,the number of pixels are counted by tracking the pixel chain of the skeleton and,finally,the fiber length is obtained by calculating ratio of the calibrated length and the actual length.The experiment shows that the method of automatically measuring fiber length in this paper has better robustness for the uncomplicated background without cross-fiber images.
作者
龚平
梁菁
Gong Ping;Liang Jing(School of Electronic and Information Engineering,Wuzhou University,Wuzhou 543002,Guangxi,China;School of Big Data and Software Engineering,Wuzhou University,Wuzhou 543002,Guangxi,China)
出处
《梧州学院学报》
2021年第3期1-7,共7页
Journal of Wuzhou University
基金
广西创新驱动发展项目(重大科技专项)(AA18118036)
广西自然科学基金项目(2018JJB170021)
关键词
纤维长度
图像处理
自动测量
边缘检测
单像素骨架
Fiber length
Image processing
Automatic measurement
Edge detection
Single-pixel skeleton
