Efficient pore space characterization based on the curvature of the distance map
Media classification and the construction of pore network models from binary images of porous media hinges on accurately characterizing the pore space. We present an efficient method for (i) locating critical points, that is, pore body and throat centers, and (ii) partitioning of the pore space using information on the curvature of the distance map (DM) of the binary image. Specifically, we use the local maxima and minima of the determinant map of the Hessian matrix of the DM to locate the center of pore bodies and throats. The locating step provides structural information on the pore system, such as pore body and throat size distributions and the mean coordination number. The partitioning step is based on the eigenvalues of the Hessian, rather than the DM, to characterize the pore space using either watershed or medial axis transforms. This strategy eliminates the common problem of saddle-induced over-partitioning shared by all traditional marker-based watershed methods and represents an efficient method to determine the skeleton of the pore space without the need for morphological reconstruction.