活塞表面缺陷图像的亚像素边缘检测技术研究  

Research on Sub-Pixel Edge Detection Technology of Piston Surface Defect Image

作　　者：郑彬 黄涛 罗山 ZHENG Bin;HUANG Tao;LUO Shan(School of Transportation and Automobile Engineering Panzhihua University,Sichuan Panzhihua 617000,China)

机构地区：[1]攀枝花学院交通与汽车工程学院,四川攀枝花617000

出　　处：《机械设计与制造》2024年第1期139-142,149,共5页Machinery Design & Manufacture

基　　金：国家自然科学基金项目(U1960101);攀枝花市指导性科技计划项目(2019ZD-N-2)。

摘　　要：针对活塞图像对比度低、缺陷区域小、缺陷种类多、人工检测效率低等问题,提出一种结合区域生长法和亚像素边缘提取的活塞表面缺陷检测方法。利用图像处理软件采集活塞图像,分析活塞表面图像中缺陷区域与正常区域灰度值的差异。使用区域生长法进行图像分割,结合Canny算子对活塞表面缺陷边缘进行初步定位。通过定位感兴趣区域的位置,进行亚像素级别提取,并平缓感兴趣区域边缘。实验表明,所提算法比传统的边缘提取方法得到的缺陷区域更精确、平滑。Aiming at the problems of low contrast,small defect area,many kinds of defects and low efficiency of manual detection in piston image,a piston surface defect detection method combining region growing method and sub-pixel edge extraction is pro-posed.The piston image was collected by image processing software,the difference of gray value between defect area and normal area in piston surface image is analyzed.The region growing method is used for image segmentation,and the Canny operator is combined to preliminarily locate the surface defect edge of the piston.By locating the location of the region of interest,sub-pixel level extraction is carried out,and the edge of the region of interest is smoothed.Experimental results show that the proposed algo-rithm is more accurate and smooth than the traditional edge extraction method.

关 键 词：区域生长法 亚像素边缘检测 感兴趣区域 图像分割 缺陷检测 

分 类 号：TH16[机械工程—机械制造及自动化]