Abstract: Actual measurement of rock samples of reservoirs is the main traditional method for obtaining their physical-property parameters, but this method can not be fit for analysis of carbonate microstructure. By means of Matlab powerful digital image processing functions, the digital images of carbonate rock samples obtained from the environmental scanning electron microscope (ESEM) are processed to get statistical pore characteristic parameters: image porosity, calculated permeability and circle-like parameters. And then, correlation analysis of these parameters with the actual measurement data of porosity and permeability is conducted. The results show that using digital image analysis method for obtaining characteristic parameters of carbonate rocks can correctly evaluate the physical properties of rocks, and get more information of the characteristic parameters than traditional methods.

Key words: carbonate rock, digital image, porosity, permeability, ESEM