Abstract: The accumulation and migration of coalbed methane in coal seams is affected by the characteristics of structure and shape of pores and fractures in coal. To characterize the type and spatial distribution of the pores and fractures in coal, the paper took the medium rank coal from Ningwu basin, used micro-CT scanning technology to scan to get CT images and adopted the FDK algorithm for construction of grayscale images, then based on experimental porosity, used binary segmentation to cut the grayscale images and recognizing the micro-fractures in CT image. By using the images obtained by micro-CT scanning technology and based on the coal matrix, pores (fractures) and mineral components showed a different distribution of CT values and finally the 3D modeing of pores and fractures of coal was realized, and the dimension and the spatial distribution of pores and fractures were characterized comprehensively. The research results have great contribution to research the process of gas adsorption-desorption and seepage in coal.