基于Otsu算法和Sobel算子的连接件图像边缘检测  

Linker Image Edge Detection Based on Otsu Algorithm and Sobel Operator

作　　者：王鲲鹏[1,2] 

机构地区：[1]常州市软件技术研究与应用重点实验室 [2]常州工学院计算机信息工程学院

出　　处：《现代计算机（中旬刊）》2011年第7期17-19,共3页Modern Computer

摘　　要：着重研究基于机器视觉的薄片连接件检测系统中的图像边缘检测方法。该方法首先采用Otsu算法对CCD工业相机获取的连接件灰度图像进行阈值分割,然后再用Sobel算子进行边缘检测。整个流程采用MatLab进行仿真测试。测试结果表明,该方法改善了单一采用Sobel算子检测的准确性,能提高边缘检测的整体性能,检测精度较好,该方法是实用可行的。Focuses on studying the image edge detection method of connector detection system based on machine vision.This method first uses Otus algorithm to make threshold segmentation on the connecting piece of gray image obtained by CCD industrial cameras,and then uses Sobel operator to make edge detection.Uses MatLab to make simulation test in the whole process.The test result shows that this method can improve the detection accuracy of single using Sobel operator and the overall performance of edge detection.This method has the higher detection accuracy and is practical and feasible.

关 键 词：边缘检测 OTSU算法 SOBEL算子 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]