In this paper we present 2-D and 3-D image-based methods to detect fractures in drill core. With the first approach, fractures are detected from a profile along the central axis of a core sample. Fracture detection strategies include taking the difference between the core profile and a model of unbroken core, identifying changes in slope with the first derivative and using the mean as a threshold. Using the mean as a threshold is the simplest and most reliable fracture detector of the three profile methods. The second approach is to convert 3-D point cloud data to a 2-D image, with each pixel representing a distance measurement. These distance-based images do not show rock texture, thus established edge and object detection tools are better able to highlight fractures. Once the fractures are identified, the length of each intact core section is measured. From these measurements the Rock Quality Designation (RQD) is calculated.

This content is only available via PDF.
You can access this article if you purchase or spend a download.