Image processing for automatic extraction of rock joint orientation data from digital images

A complete algorithm for starting with a digital image of a rock face, automatically delineating traces, and extracting three dimensional joint or fracture orientation data is described. Less emphasis is placed on the pre-processing of the image, as different images require different levels of filtering, but nearly all pre-processing ultimately produces a binary image resembling a trace map. The special R-Theta Hough transform is then used to find and extract linear features from the binary image. Using relationships derived from vector calculus, and assuming a Fisher distribution for each structure set, a forward Monte Carlo simulation is performed to determine mean dip and dip direction of the set, as well as the Fisher constant, K. A case study is presented to illustrate the performance of the algorithm, however more case studies are needed to compare results gained with image processing to results of a scanline survey.