摘要
在纱线质量测试过程中,要求对纱线的外观进行检测以保证纱线的品质可靠。找到纱线中存在的不合格特征,能有效地提高纱线的生产效率与质量。根据纱线图像特点,分析出背景对纱线的影响,对单行纱线图像的灰度图进行分析,提出了基于双峰法的阈值分割法。通过对比不同阈值处理后的纱线图像,表明当阈值为纱线图像最大灰度值的10%~35%时,图像分割结果稳定,阈值为最大灰度值的30%时,具有最佳的阈值分割效果。由于阈值运算简单,可以满足测试系统要求,也为纱线图像检测智能化提供了可靠应用。
During yarn quality testing, the appearance of yarn should be tested to ensure reliable quality. Finding out the unqualified features in yarn can effectively improve the production efficiency and quality of yarn. According to the characteristics of the yarn image, the influence of the background on the yarn is an- alyzed. The grayscale image of the single-row yarn image is analyzed, and the threshold segmentation method based on the bimodal method is proposed.By comparing the yarn images processed by different thresholds, it is shown that when the threshold value is 10%-35% of the maximum gray value of the yarn image, the image segmentation result is stable, and the threshold value is 30% of the maximum gray val- ue, which has the optimal threshold segmentation effect.Because the threshold calculation is simple, it can meet the speed requirement of the test system, and it also provides a reliable application for the intelligent detection of yarn image.
作者
张增康
马卫红
ZHANG Zeng-kang;MA Wei-hong(School of Optoelectronic Engineering,Xi＇an University-of Technology,Xi＇an 710021,China)
出处
《化纤与纺织技术》
2018年第3期34-38,共5页
Chemical Fiber & Textile Technology
基金
陕西省科学技术研究发展计划项目(2014K05-41)
陕西省教育厅重点实验室科研计划项目(15JS036)
关键词
纱线
双峰法
阈值
图像分析
yarn
bimodal method
threshold
image analysis
作者简介
张增康（1993-），男，陕西周至人，硕士，实验员，从事机器视觉检测、图像处理研究。;通讯作者：马卫红（1976-），女，陕西大荔人，博士，副教授，从事光电测试、图像处理研究。
