一种基于图像处理技术的柴油机单颗粒微观结构的分析方法  

A Methodology for Microstructure Analysis of Diesel Engine Particulate Based on Image Processing

作　　者：宋崇林[1] 张炜[1] 王林[1] 吕刚[1] 李方成[1] 

机构地区：[1]天津大学内燃机燃烧学国家重点实验室,天津300072

出　　处：《燃烧科学与技术》2010年第5期388-395,共8页Journal of Combustion Science and Technology

基　　金：国家自然科学基金资助项目(51076115)

摘　　要：为了准确地定量分析柴油机单颗粒微观结构,采用Gabor滤波法、局部阀值法、改进的OPTA法等图像处理技术对微粒微观形貌图像进行特征提取,获得了微粒微观结构特征参数微晶尺寸、层面间距、曲率的计算统计方法.试验结果表明,该方法可以快速、准确地获取微粒微观结构特征参数,偏差率仅为0.15%,是研究柴油机微粒微观结构的重要工具.试验结果还表明,随着柴油机负荷的增加,微晶尺寸平均值从2.161.nm增加到2.180 nm,而层面间距平均值和曲率平均值则分别从0.362.nm和1.228减小到0.360.nm和1.224.Image processing algorithms,including Gabor filter,local threshold method and improved OPTA algorithm,have been developed to extract microstructure characteristics of diesel particulate,such as fringe length,tortuosity and separation from high resolution transmission electron microscopy images.By this method,the particulate microstructure characteristics can be analyzed accurately and rapidly.The experimental data showed that the deviation was only 0.15%,thus proving the method to be an important tool for studing diesel particulate microstruture.To illustrate the utility of the analysis parameters for diesel particulate nanostructures,two particulates sampled under different operating conditions were used.The preliminary experimental results indicated that with the increase of load,mean fringe separation distance and mean tortuosity decreased from 0.362 nm and 1.228 to 0.360 nm and 1.224,respectively,which mean fringe length increased from 2.161 nm to 2.180 nm.

关 键 词：柴油机颗粒物 微观结构 图像处理 微晶尺寸 层面间距 曲率 

分 类 号：TK421.5[动力工程及工程热物理—动力机械及工程]