The Research Of Screen-space Ambient Occlusion Algorithm On Protein Surface Model Rending

As an important assistant method on biological molecular research,modeling and visual analysis technology provides intuitionistic observe method and abundant real-time data for researchers relying on powerful data processing and analysis capacity of computers.However,the development of biological research and medical analysis makes the size of protein dataset increase to ten thousand magnitude or even hundred thousand magnitude,which makes a challenge for dynamic real-time processing.The way to realize real-time processing of macromolecule using finite sources becomes the hot and difficult spot in biological and computer scientific fields.This paper tries to combine a popular technology called screen-space visualization with solvent exclusion surface(SES)model of molecules to realize visualized effect on protein.Using random point sampling methods depending on depth surface to collect occlusion information around the inspection point,analyzing the drawbacks of the current sampling algorithm and puting up with improved schemes;relocating sampling positions by means of calculating intersection points between light direction and depth space to improve validity of the data and vraisemblance of the rending result.Specific studies are as follows:To begin with,we come up with two different samping algorithms to improve the veracity of current screen-space-based sampling algorithm.Improved algorithm designs consider both Monte-Carlo random principle and depth feature of pixels,which make sampling results reflect furthest molecular real surface occlusion characters.And then,we noticed that some sampling positions in screen space are erroneous,so a relocation method of sampling points is proposed.Through collecting all the intersection points between lights casted from inspection point and pixel depth space,we find a nearer occlusion position to the inspection point,by which enhance the validity of ambient light calculate result.Finally,the visualization program framework used in this research is introduced,and the proposed methods are verified with experiments and analyses to prove the effectiveness and instantaneity.