Medição da elevação da superfície d’água em experimentos 2d usando análise de imagens: uma ferramenta de validação na engenharia naval e oceânica

Abstract

In Naval and Ocean Engineering, several complex phenomena of fluid-structure interactions are commonly evaluated through simplified two-dimensional hydrodynamic experiments. The measurements obtained through these experiments, such as the spatial and temporal evolution of the water surface elevation, are of great importance for the validation of analytical and numerical models. With the constant evolution of optical devices and the continuous decrease in their cost, image-based measurements have been applied to quantify water surface movement in experiments, using “virtual” sensors. However, there may still be errors in the final measurement, caused by different types of distortions in the images used. The study of these errors in measurements is relevant to determine the level of reliability of data generated from images. The present work consists in the evaluation of errors in measurements, based on images, of time series of water surface elevation in two-dimensional hydrodynamic tests. For this, videos of hydrodynamic experiments available in the literature were used, related to the problem of shipping water on the deck of a fixed structure. Time series of water elevations at different positions were obtained from these videos, using open-source image processing and analysis techniques. The influence of varying image distortions on water surface elevation measurements was evaluated, including variation in azimuthal projection, focus, noise, and scale. The results showed that the effects due to scale and noise distortions were the most considerable. On the other hand, the distortions due to the azimuthal and focus distortions showed less differences in relation to the measurements of the reference cases. This type of studies can contribute to research that uses measurements based on image analysis, as it is recommended to express measurements with an associated error level.