Abstract
The deterioration of concrete structures is a growing problem worldwide. Drilled cores are usually collected from large dams for testing the concrete and rock foundation underneath to assess their safety. However, the borehole itself can be used to assess cracks and other damage and collect additional information on the surrounding materials. This paper evaluates various edge-detection algorithms, as well as transform and statistical-based methods, for their effectiveness in assessing damage in a concrete dam from digital borehole imagery obtained using an acoustic televiewer. The statistical-based approach was found to be the most efficient technique for damage assessment from acoustic imagery. A clustering technique was used to quantify damage from the imagery, such as vertical cracks, horizontal cracks, voids, stains, and foundation damage. Results were verified using log data. Further damage analysis consisted of determining minimum, maximum and mean crack-width openings.