低对比度工况下轴安装间隙提取方法研究  

Research of Shaft Installation Clearance Extraction under Low Contrast Condition

作　　者：甘玉坤 丁跃浇[1] 罗振威 吴雪婷 李旭东 GAN Yu-kun;DING Yue-jiao;LUO Zhen-wei;WU Xue-ting;Li Xu-dong(Hunan Institute of Science and Technology,Yueyang 414000,China)

机构地区：[1]湖南理工学院,湖南岳阳414000

出　　处：《计算机技术与发展》2020年第8期183-187,共5页Computer Technology and Development

基　　金：国家重点研发计划“制造基础技术与关键部件”专项(2018YFB2001400)。

摘　　要：针对轴安装间隙图像在间隙处对比度较低的问题,提出一种通过分段线性函数和图像反转来增强图像边缘信息的方法。通过计算图像直方图以确定安装间隙图像的灰度值范围,利用安装间隙的灰度值范围确定线性函数的分段点。用分段线性函数分别处理图像中安装间隙与间隙外的其他区域,将安装间隙灰度值区间之外的灰度值设置为固定值,去除其他区间的灰度值对间隙提取的干扰,同时拉伸间隙灰度区间内的图像对比度,提升间隙处图像的对比度。通过图像反转来反转进行图像增强之后的图像,反转间隙与其他区域的灰度值后,使用Canny算法提取间隙边缘。实验结果表明:该方法能去除图像中安装间隙以外的区域对间隙内图像的影响,提升Canny算法对轴安装间隙边缘信息的提取效果。Aiming at the problem that the contrast to the axial installation gap image is low in the gap,a method of enhancing the edge information about the image by piecewise linear function and image inversion is proposed.The segmentation point of the linear function is determined by calculating the image histogram to determine the range of gray values of the installed gap image,using the range of gray values of the mounting gap.The piecewise linear function is used to separately process the installation gap and other areas outside the gap,and the gray value outside the gray interval of the installation gap are set to a fixed value,and the interference in the gray value of the other interval of the gap extraction is removed,and the interference is pulled at the same time.Extend the contrast to the image of the grayscale interval of the gap to increase the contrast to the image of the gap.The image after the image enhancement is reversed by image inversion,and the gray value of the gap and other regions are inverted,and the gap edge is extracted by the Canny algorithm.The experiment shows that the proposed method can remove the influence of the area outside the installation gap between the image of the gap,and improve the Canny algorithm’s extraction effects on the edge of the shaft installation gap.

关 键 词：安装间隙 直方图 分段线性函数 图像反转 边缘提取 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]