Image processing, filtering and segmentation of data-sets for reservoir simulation, 2013
Tiril Pedersen Gurholt
Advisors: Ivar Aavatsmark
Short description of project:
This thesis is organized into three parts. In Part I the objective is to improve the presentation of 2D image time-series of fluid flow experiments acquired using an in-house X-ray scanner. This scanner introduces severe horizontal and vertical line artifacts which partly occlude the depicted fluid flow. These artifacts can be attenuated in the frequency domain after constructing an appropriate filter. In this part an introduction to frequency domain filtering and the filter construction is iv given, before presenting the results and giving a conclusion. In Part II the objective is to study the performance of various image segmentation techniques for segmentation of CT images of vuggy carbonate rocks. These image volumes have varying image resolution and therefore also varying degree of difficulty. In this part an introduction to the segmentation techniques are given, before presenting the results and giving a conclusion. In Part III the included papers are presented.
Link to thesis at BORA-UiB: Link