砂土颗粒SEM图像分析方法的研究

A Study of Sandy Soil Granule Analysis Method Based on SEM Image

提取砂土颗粒SEM图像的细观特性参数对砂土颗粒的细观研究具有极其重要的意义。在提取砂土颗粒的细观特性参数时遇到了图像质量不高,粘连颗粒比较严重等问题,影响了特征参数的提取。针对这个问题,对砂土颗粒SEM图像的特征进行了分析,借助开运算、中值滤波、直方图均衡化等方法,对图像进行处理,优化了图像质量。使用Otsu法二值化,对处理后的砂土颗粒图像应用形态学处理,改进的分水岭分割算法等,分割粘连的砂土颗粒。最后提取了砂土颗粒的数目、占空比、质心等参数,为岩土的组构张量描述提供了数据支撑。

The study analyzed the sandy soil granule SEM image feature microscopic characteristics parameter have extremely important significance. In the extraction sandy soil granule feature microscopic characteristics parameter of poor image quality, serious particle adhesion effects of extraction of the feature parameter. To solve this problem, the study analyzed the sandy soil granule SEM image first and put forward the methods of the open operation, smooth operation, histogram equalization on image processing to achieve image enhancement. The study also used Otsu method on image of processed sandy soil granule to have morphology processing and improved watershed algorithm method, ete.to partition parliele adhesion. The final part extracting sandy soil granule＇s quantity, duty cycle and eentroid, etc . Provides data support for the tensor description of geotechnical fabric.