Microstructure identification and statistical method of the soil SEM image

Qiang TANG,Chun LIU,Ying-fan GU,Bin SHI

DOI: https://doi.org/10.3969/j.issn.1674-9057.2017.03.025

2017-01-01

Abstract:Based on the image processing software PCAS,this paper studies the quantitative analysis method of the soil microstructure.By explaining the significance of threshold (T),pore throat closed radius (r),minimum pore area (S0) and statistical parameters during image identification.The method of selecting these parameters (T,r and S0) is proposed when the soil SEM image is processed.The SEM images of six clay samples were processed bysoftware,with different r and S0.Not only can the pore distribution image of the clay be analyzed automatically,but also the geometrical parameters and statistical parameters of the pore structure are obtained accurately,and the quantitative analysis of the pore structure in the clay is realized.The results of the image analysis finally show that the result image the pore distribution can better reflect the microstructure of the soil,with smaller r and S0.When there are less apparent porosity and the noise of the binary image.The microstructure image can be analyzed by increased r and S0 appropriately when there are more apparent porosity of the soil and noise in the binary image.