应用于电缆绝缘体检测中的图像分割技术的研究  

Research of image-segment technique in cable-insulator testing

作　　者：周明翰[1] 江斌[1] 陈璐玲[1] 李孟超[1] 

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093

出　　处：《光学仪器》2009年第6期16-20,共5页Optical Instruments

摘　　要：应用数字图像处理技术来测试电缆绝缘体的参数,图像分割是这一数字图像处理技术中最重要的部分。针对电缆绝缘体截断面图像的特征,利用图像分割技术中的改进的全局门限、高斯滤波和Sobel算子等算法,实现了电缆绝缘体图像的分割和提取,有效的消除了阈值分割所产生的边缘锯齿。仿真和实际运行结果表明,该方法检测精度高,有效提高了检测速度,能够实现在工业上的大规模应用。To test the parameters of the cable insulation with application of digital image processing techniques, of which image-segment technique is the most important as a part. In this paper, we use the arithmetic of advanced global threshold, Gaussian filtering and sobel arithmetic operators etc to realize that the segmentation and extraction of cable insulator image in image-segment technique, which effectively eliminate the serrated edge on threshold segmentation, according to image-feature of cable insulation. The results of simulating and practical application show that the method of detection has high precision, and effectively improve the detection rate, which can achieve the large-scale industrial application.

关 键 词：图像分割 改进的全局门限 高斯滤波 SOBEL算子 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]