Image Interpolation Technology Research And In Medical Volume Data Visualization Applications

Volume data sets visualization is applied in medical domin widely for recent years. And Image interpolation is of great importance in biomedical visualization and analysis.Image interpolation is usually required for proper visualization, as the 3D data sets must be isotropic in order to produce the correct aspect ratio along each direction when displayed. Broadly, interpolation techniques can be divided into two categories: scene-based and object-based. The main research and contributions from this paper include:In scene-based methods, interpolated scene density values are determined directly from the density values of the given scene. Because of the unrealization of sin c kernel in the course of reconstruct,we usually substitute it using approximate function.First,we can use window function,second we can using polynomial.We also analyse the performance of this kernel both in time and frequency region.This kind of method is widely used in practice because of its high speed and efficiency.But the blur of boundary is unavoidable because of its low pass performance.In object-based methods, some object information extracted from the given scene is used in guiding the interpolation process.The method of contour change is used in binary image commonly.But now gray image appears more and more.The traditional method based on gray-level is suitable for the slices of small distance and shape difference .Otherwise , blur of the object's boundary will appear.In order to solve the two question just mentioned,this thesis proposed a method which uses a polygon to approximate the object shape and performs the interpolation using polygon as references,which combine the shape information and gray-level information together .The experiment results attests the validity.