基于GLCM的磨粉机喷砂辊表面图像纹理特征分析  

Analysis of Texture Features of Surface Image of Sandblasting Roller of Mill Based on GLCM

作　　者：刘培康 武文斌 王雪峰[2] 侯宁沛 张文龙 吕少杰 赵歧峰 LIU Peikang;WU Wenbin;WANG Xuefeng;HOU Ningpei;ZHANG Wenlong;LV Shaojie;ZHAO Qifeng(Grains&Oils Machine Institute,Henan University of Technology,450001 Zhengzhou,China;Zhengzhou University of Science and Technology,450001 Zhengzhou,China)

机构地区：[1]河南工业大学粮油机械研究所,郑州450001 [2]郑州科技学院,郑州450001

出　　处：《粮食加工》2022年第5期11-15,共5页Grain Processing

基　　金：河南省重点研发与推广项目(小麦磨粉机智能磨辊的研究开发:2018304)。

摘　　要：为了定量表征磨粉机喷砂辊表面形貌、准确判断喷砂辊表面磨损状态,对河南麦道面粉有限公司2M工艺部位DMFT型磨粉机喷砂辊,采用搭建的图像采集装置对磨辊磨损周期内的表面形貌进行采集;基于灰度共生矩阵算法,并应用matlab平台进行图像纹理特征提取。研究了4种常用的纹理特征参数随磨损时间的变化规律,结果表明:随着喷砂辊表面磨损时间增加,其表面形貌纹理特征参数二阶矩、熵值和对比度均呈现单调递减的变化,该图像纹理特征及变化规律对区分喷砂辊表面磨损程度有良好的体现,可以用于定量表征喷砂辊形貌,为准确描述喷砂辊表面磨损状态提供参考。In order to quantitatively characterize the surface morphology of sandblasting and accurately judge the surface wear state of sandblasting roller.According to the sandblasting roller of DMFT mill in 2M technological position of Henan McDonnell Flour Co.,Ltd.,the surface morphology in the wear period was collected by the built image acquisition device.Based on the gray level co-occurrence matrix algorithm,the image texture features were extracted by using matlab platform,and the variation rules of four commonly used texture feature parameters with wear time were studied.The results show that the second moment,entropy and contrast of the texture characteristic parameters of the sandblasting roller decrease monotonously with the increase of the surface wear time,which can be used to quantitatively characterize the sandblasting roller,and provide a reference for accurately describing the surface wear state of the sandblasting roller.

关 键 词：灰度共生矩阵 喷砂辊 磨损 表面图像 纹理 特征提取 

分 类 号：TS211.3[轻工技术与工程—粮食、油脂及植物蛋白工程]