Pores of rocks are important places to store oil and gas in reservoirs. Analyses of pore structure by cast thin sections are of great significance to reservoir quality evaluation. However, pore structures of carbonate rocks are relatively complex, and there are a lot of noises and interference factors in the image of cast thin sections, which leads to the poor effect of porosity extraction by conventional methods.Methods:A pore extraction method from cast thin- sections by multiple thresholds (ctsPore method) was introduced in this paper to extract the pore and estimate the plane porosity. By comprehensively utilizing different physical property parameters in HSV color space, this method realizes accurate extraction of pore area and solves the problem that the dissolution pore is easily confused with noises.Results:①In this paper, a set of threshold parameters of ctsPore method is determined according to the characteristics of carbonate reservoirs: H is 180°~240°, dt is 0.12, T1 and T2 are 320 mm2. The average absolute error and relative error of this method and PSQ method are 0.24% and 5%; ②It only takes about 2.5 seconds to extract a 2560×1920 resolution image, which can be used to quickly extract pores of a large number of cast thin- sections images and provide feasible technical support for analyses of reservoir pores; ③The ctsPore method is used to automatically extract pores from cast thin- sections of A Oilfield in Iraq. The results show that the porosity of the study area is mainly between 6% and 8%, and the reservoir is a relatively tight—low porosity reservoir; ④The porosities of A1~A3 layers are 6.9%, 11.7% and 5.4% respectively, which are consistent with the vertical distribution of physical properties, sedimentary microfacies and lithology of carbonate reservoirs in the study area.Conclusions:In this paper, a set of threshold parameters of ctsPore method are determined according to the characteristics of carbonate reservoirs. The experiments of accuracy and efficiency analyses show that the porosity results of carbonate cast thin sections based on ctsPore method are consistent with the real situation. The porosity extraction results are consistent with vertical characteristics of well profile, including reservoir physical properties, sedimentary microfacies and lithology.