臺灣博碩士論文加值系統

在這篇論文中,我們研究3D醫學影像的視覺化。無論是CT或MRI影像中,我
們認為必須應用影像處理的技術,以其達到較高的自動化,進而提高立體資
料視覺化的用途。在我們的實驗結果顯示,在影像處理之後,視覺化的方法
要做適當的調整,最主要是surface normal的訂定,我們利用影像處理的技
術處理以上的問題。在實驗的過程中,我們發現CT取像通常都不會太多層,
在resolution不夠時,任何視覺化的方法都難以做出正確的、等比例的
reconstruct。在此論文中,我們用Morphological interpolation應用著
影像處理的技術來解決上述的問題。

This paper deals with the research of volume rendering of 3D
medical data sets. Our goal is to volume render the CT and MRI
data sers. We believe that we must apply the technology of
image processing to obtain more automation and increase the use
of volume rendering. Based on the experimental results, the
volume rendering is done a proper adjustment after applying the
image processing. It is general the surface normal. We adopt
the technology of image processing (Sobel operator) to do and
boundary extraction and get the surface normal. In our
experimental process, we discover that the CT slices are not
too much. These CT data sets have low resolution in z direction
than in x or y directions. As low resolution, any volume can
not do the correct and equal ratio reconstruction. In this
paper, we apply morphological interpolation and the technology
of image processing to slove above problem.