Accurate and efficient maximal ball algorithm for pore network extraction. (April 2017)

• Record Type: Journal Article
• Authors: Arand, Frederick; Hesser, Jürgen
• Is Part Of: Computers & geosciences. Volume 101(2017)
• Page Start: 28
• Page End: 37
• Publication Date: 2017-04
• Subjects: Pore-Throat-Network -- MIS -- Mathematical morphology; Environmental policy -- Periodicals; 550.5
• DOI: 10.1016/j.cageo.2017.01.004
• Languages: English
• ISSNs: 0098-3004

Abstract

Abstract: The maximal ball (MB) algorithm is a well established method for the morphological analysis of porous media. It extracts a network of pores and throats from volumetric data. This paper describes structural modifications to the algorithm, while the basic concepts are preserved. Substantial improvements to accuracy and efficiency are achieved as follows: First, all calculations are performed on a subvoxel accurate distance field, and no approximations to discretize balls are made. Second, data structures are simplified to keep memory usage low and improve algorithmic speed. Third, small and reasonable adjustments increase speed significantly. In volumes with high porosity, memory usage is improved compared to classic MB algorithms. Furthermore, processing is accelerated more than three times. Finally, the modified MB algorithm is verified by extracting several network properties from reference as well as real data sets. Runtimes are measured and compared to literature. Abstract : Highlights: Subvoxel precise pore network extraction with modified maximal ball algorithm. Morphology volume improves algorithmic speed and memory efficiency. Benchmarking of modified algorithm with artificial and real data sets. Comparison to other algorithms found in literature.