基于Woodcock Tracking的快速蒙特卡罗直接体绘制方法  

Woodcock Tracking Based Fast Monte Carlo Direct Volume Rendering Method

作　　者：周守东[1] 

机构地区：[1]安徽工业经济职业技术学院,安徽合肥230051

出　　处：《系统仿真学报》2017年第5期1125-1131,共7页Journal of System Simulation

基　　金：安徽省重点自然科学研究项目(KJ2015A381);安徽省高校优秀青年人才支持计划(gxyq2017177)

摘　　要：Woodcock-Tracking采样算法是一种常用的光线采样方法,对于优化蒙特卡罗直接体绘制方法具有重要的作用。目前该方法没有考虑体数据不同区域颜色的变化,导致颜色变化剧烈的区域可能只有少量采样点,影响体数据的真实性。引入信息熵表征体数据不同区域变化的剧烈程度,基于信息熵发展了自适应Woodcock-Tracking方法,使得变化剧烈区域获取更多采样点;运用加权颜色合成算法,实现无偏渲染。通过对头颅、背包和骨骼三个模型的对比测试,结果均表明该方法对渲染质量有明显改善。Woodcock-Tracking sampling algorithm is a frequently applied ray sampling method, which plays an important role in optimizing the Monte Carlo direct volume rendering method. However, without considering the volume of color＇s variation in different region by using this method, the situation that region with great color＇s variation may only get a few samples was caused and the authenticity of volume data would be affected. Entropy to indicate color＇s variation in volume was introduced, an entropy based adaptive Woodcock-Tracking method was developed, which could sample more points in region with higher color＇s variation. To overcome biases caused by adaptive Woodcock-Tracking method, weighted color evaluation method was used, so that the final render result image was unbiased. Through the contrast test on the three models of head, backpack and skeleton, the result demonstrates the presented method is superior to the current method in the aspect of improving the rendering quality obviously.

关 键 词：直接体绘制 全局光照 蒙特卡罗光线跟踪 信息熵 Woodcock-Tracking 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]