Abstract: In order to characterize the fracture aperture and spatial distribution of the completed coal sample, the high precision Micro-CT scanning technology was applied to analyze th e crack structures of the coal samples from Xinjulong Mine. According to the CT number distribution of the crack surrounding in the CT image and 3D visualization technology, a fine characterization was conducted on the coal and rock crack apertures, distribution, space allocation and the parameters. The study result showed that the application of the Gauss function could overcome the CT numbers along the two sides of the cracks, due to the fluctuation caused by the heterogeneity and could well fit the attenuation law of the crack CT number curve. The kurtosis of the Gauss function and the curve area were applied to measure the crack distribution shape. When the crack aperture value was small ( 0.24 mm) , the measurement with the MA method would have high accuracy. The crack 3D visualized model could be reconstructed with the coal and rock 3D and could clearly recognize the crack space distribution shape in the coal. According to 197 feature locations in the 3D crack area, the error analysis of the CT number showed that the average error measured with the parameter PH measurements were 14 .3%. The average error measured with the MA method measurement was 17 .1%. The study results showed that PH method has higher accuracy.