The effectiveness of different thresholding techniques in segmenting micro CT images of porous carbonates to estimate porosity

Abstract The effectiveness of various thresholding techniques in segmenting micro X-ray computed tomography (XCT) images of porous carbonates has been investigated using experimental analysis. A comparison between directly measured and image-derived porosities clearly exhibited that the application of different segmentation methods produced vastly different results. The obtained results demonstrated the importance of the segmentation step for quantitative pore space analysis. In this research, three global thresholding methods and three locally adaptive thresholding methods were examined and only a few of the tested methods yielded acceptable results. Generally, locally adaptive methods generated better results; however, as the performance of global methods were reduced because of presence of anhydrite concentrations in the studied samples, a pre-segmentation for determining the anhydrite phase increased the performance of GT methods dramatically. As a result, two-threshold Otsu method yielded the best outcome. The novelty of this work is to compare the performance of different methods in segmentation of XCT images and to focus on the carbonates of one of the southern oilfields of Iran.