In order to develop formulae to forecast the damage evolution of rubble mound breakwater armour layers, long duration scale model tests have to be carried out. To help in the operation of these scale model tests, a procedure for stereo image reconstruction of a submerged scene was developed. The accuracy of the tri-dimensional reconstruction of the scene captured in the stereo pair depends on the resolution of the images in the stereo pair. However, increasing the image resolution has a significant impact on the computational effort needed. To determine which one provides an accurate reconstruction, several stereo pair reconstructions were carried out using different image resolutions.
The design process of the armour layer of rubble-mound breakwaters is still at a pretty rudimentary stage. The safety level adopted is determined solely by the choice of the return period of the characteristic wave height of a sea state that is deemed to better describe the wave climate at the structure location. In addition, being the armour layer of rubble-mound breakwaters a component whose failure is noticeably progressive, it is not reasonable to design it for a single sea state: The first armour layer damage occurs well before its failure and it is common for this damage progression to take several years. Taking this into account, the development of formulae to forecast damage evolution of rubble-mound breakwater armour layers is of paramount importance. In fact, although there is already some work in this domain of armour layer damage evolution forecast, the available formulae (Melby and Kobayashi, 1999) were obtained from a rather limited set of long duration scale model tests with rubble-mound breakwaters whose armour layer was made of rocks, which is not the most common armour layer unit in the structures of this kind that are built in Portugal.
