Versatile and efficient pore network extraction method using marker-based watershed segmentation

摘要: Obtaining structural information from tomographic images of porous materials is a critical component media research. Extracting pore networks particularly valuable since it enables network modeling simulations which can be useful for host tasks predicting transport properties to simulating performance entire devices. This work reports an efficient algorithm extracting using only standard image analysis techniques. The was applied several ranging sandstone fibrous mats, and in all cases agreed very well with established or known values throat sizes, capillary pressure curves, permeability. In the case sandstone, present compared obtained current state-of-the-art algorithm, good agreement achieved. Most importantly, extracted correctly predicted anisotropic permeability tensor, demonstrating ability detect key features. highly allows extraction on fairly large 500^{3} voxels just over 200 s. one match as varied 20% porosity 75% significant advancement. source code this provided.