Computerized X-ray micro tomography (micro CT) is a powerful technique for studying the structure and properties of porous scaffold. A variety of sample parameters can be studied using a single micro CT scan . In general a segmented data set is a required for most quantitative analysis. However, segmentation of CT data can be difficult due to the artifacts in micro CT images such as blurred interfaces due to the machine contrast transfer function and the partial volume effect. Therefore the segmentation can be biased and prone to errors. Many methods have been developed to improve segmentation, however the interface problem has not been solved perfectly . Porous scaffolds suffer from these effects because of their high surface to volume ratio and hence large interface. In this paper we discuss the interface problem in detail and demonstrate the effect of voxel size on the histograms of CT images of porous scaffold as well as a thresholding method based on 2 dimensional histogram is also presented. The potential of this method in more complicated scenarios such as 3-phase system is currently being investigated .