Gray-scale Image Segmentation And Pore Morphology Research In Shale

Now,shale gas plays an important role in solving the energy problem.By Studying the pore-scale properties of the shale,we can explain the fluid mechanism inside of the pores fundamentally.Before analyzing the shale on the micro-scale,we should use the slice images to construct the pore network.We should attain binary image of the shale first,then we can analysis the pore structure using the statistics data.Shale possesses low porosity(much below 20%),its average porosity is 5%.Pores in shale are widespread and nano scale,the structure is comparatively complex.The pore exists in both organic matter and solid,but the density of solid particles and organic matter are significantly different.As a consequence,there has no applicable algorithm for pore segmentation at present.Combined with the properties of pore and Paul’s method(we call it triangle algorithm),we proposed a new method.The main idea of this method is:for a unimodal waves,a line is drawn from the peak to the valley of the histogram.The threshold point is selected as the histogram that maximizes the perpendicular distance between the line and the point on the histogram.We compared this method with the conventional methods(otsu and shanbhag).Besides,we analysis whether the method is fit for the sandstone.At last,we study the pore-size distribution and microstructure characteristics.As a validation,this paper takes a large variety of SEM/FIB – SEM gray level image as the research object.By applying the new algorithm to the shale sample,we obtain the binary images of the sample.Besides,we obtain the pore,solid particles and organic matter separately.The results show that the algorithm is stable and efficient,this method can be applied to the image segmentation of different kinds of shale images.Furthermore,many factors influence whether the method is fit for the sandstone,such as the kurtosis value.