Fig.1 Direct Steam Injection.
Keywords: image processing, segmentation, uncertainty visualization, graph drawing, clustering.
Project Description
The ‘Direct Steam Injection’ research is focused on the turbulent mixing and heating phenomena induced by a condensing steam jet that is injected in a cross-flow of water, see Fig. 1. One of the aspects that need to be investigated is the rapid condensation of the steam in the direct neighborhood of the injection point. This can be achieved by visualizing the condensation region with a high speed camera. In order to obtain quantitative data, like the steam bubble size and its growth rate, from these recordings, it is necessary to develop a sophisticated image processing tool that enables the tracking of the gas-liquid (steam-water) interface for a sequence of images.
|
|
|
Fig.1 Direct Steam Injection. |
The aim of the project is to help automating such analysis, using image processing techniques combined with visualization tools. The first task is to segment (i.e. extract) and track the condensation region. Once this task is completed and its properties are computed, one can appropriately visualize the results to assess the process. Uncertainty visualization can also play an important role in the final visualization to emphasize the confidence level due to certain (prone-to-errors) image processing steps (e.g. segmentation). Clustering techniques and/or graph drawing can be useful to help in gaining insights about the process.
Supervisors: Andrei Jalba and Michel Westenberg