Rock joint detection from borehole imaging logs based on grey-level co-occurrence matrix and Canny edge detector

Abstract

Rock joints play an important role in characterizing the rock mass quality for geomechanical design and stability analysis. An approach was developed to detect and characterize rock joints from images collected by a borehole imaging system. A grey-level co-occurrence matrix was employed to locate the joint regions, allowing more focused and effective detection processing, followed by extractions of the upper and lower edges of rock joints using the Canny algorithm. Four basic geometric parameters of rock joints (orientation, depth, aperture and core length) were determined based on the fitting of sinusoids to joint edges. Furthermore, the joint density was determined based on the geometric parameters. To calibrate the proposed approach, a borehole on the site of the planned Rumei hydropower station at Lantsang River was selected as a case study. Orientation of rock joints with gentle dip angles, which was determined from borehole imaging logs, corresponded to the measurements in three horizontal tunnels. Additionally, both joint density and pressure-wave velocity revealed that jointed rock mass was observed at depths from 100 to 120 m, and intact rock mass was present at depths of 150–170 m, indicating the good performance of the proposed method.