Research On Intuitive & Effective Volume Data Visualization Method

With the development of information technology, the amount of 3D volume data obtained from either scientific computing or sensing device collection has grown significantly, which has resulted in an increased need to volume data visualization, so intuitive and efficient volume data visualization is becoming a focus of current research. Enhanced visual perception of interesting features in the resulting images enable users to observe and analyze volume data more directly and quickly; Accelerated rendering efficiency can ensure user-friendly interaction which is also helpful for volume data analysis and processing.. Both visual perception enhanced volume rendering and accelerated volume rendering are effective way to achieve intuitive and efficient volume data visualization. This dissertation focuses on visual perception enhanced volume rendering and accelerated volume rendering, which are effective way to achieve efficient & intuitive volume data visualization.(1) Visual perception enhanced maximum intensity projection algorithm:It is to solve the problem that Maximum Intensity Projection cann’t convey occlusion cue correctly. Firstly, K-Means clustering algorithm is used to extract structure feature automatically from result images of maximum intensity projection; and then it adjusts rendering process adaptivly according to structure features. Furthermore, the concept of rendering priority is introduced to avoid important features being occluded badly by non-important features, and an algoritm is provided to set rendering priority of structure feature automatically.(2) Faster iso-surface ray tracing by combining the K-D tree and Shell:The method takes advantage of K-D tree and Shell data structure to realize faster iso-surface ray tracing. Firstly, K-D tree is established by division of volume data space; and then Shell data is build on GPU according to K-D tree and iso-surface value; finally, faster ray tracing is realized according to search the Shell data and K-D tree in the process of iso-surface ray tracing. Furthermore, by updating Shell data dynamically in the process of ray tracing, it can avoid reconstruction and reload to GPU of Shell data when iso-surface is changing.