Coastal flooding events have caused many issues to infrastructure including bridges and highways. How to assess the flooding level and infrastructure damages in a low-cost, rapid, and accurate approach is critical to the infrastructure performance recovery. Due to the limited access to infrastructure during the post-flooding events, it is very challenging to evaluate infrastructure conditions closely. With the help of small unmanned aerial vehicles and onboard cameras, it provides the possibility to inspect the infrastructure conditions from images captured by drones remotely. With the additional help of image processing algorithms, it can help capture the infrastructure conditions and flooding levels from the imageries automatically with post-processing analysis. In this paper, we apply several different image processing algorithms to assess the infrastructure conditions by segmenting the flooding zone from the infrastructure. The performance of these algorithms in assessing infrastructure conditions is compared based on different factors with previously taken airborne imageries of infrastructure and flooding events. The performance of image processing is summarized and future work of assessing the infrastructure post-flooding damages is discussed.
|