Rock fissure pattern characterization by combining 1-D fractal dimension and statistical analysis

Abstract

How to characterize rock fissures/fractures is significant for the measurement and analysis in a lot rock engineering applications. A new method for characterizing rock fissure patterns is studied in this paper. It is constructed by combining 1-D fractal dimension and statistical analysis for the whole rock surface in an image. In a binary fissure image, to characterize fissures accurately, all the possible fissures are skeletonized, and then the short lines or curves are removed and the gaps on the main fissure segments are linking up. After that, 1-D fractal dimension is applied for analyzing fissure patterns, in which, the fissure network complexity and the fissure quantity is characterized by statistical analysis. The testing fissure images were taken from a Swedish laboratory and a Chinese Mine, and the testing results prove that the fissure pattern characterization by the new method is promising.