摘要
传输函数是实现体数据特征分类与可视化的重要手段,随着维度的升高,传输函数设计过程趋于复杂且难以实时交互;而在基于视线方向上特征分析的体数据特征可视化方法中,复杂的特征分析过程占用GPU硬件资源,降低了体绘制效率.为此提出一种快速的、不完全依赖于传输函数设计的体数据特征增强可视化算法.利用视线方向上采样点标量值变化率,快速识别视线方向上的重要特征,通过降低不重要特征的不透明度以增强重要特征在绘制结果图像中的可见性;引入反锐化掩模技术对视线方向上重要特征进行加权处理,在保证体绘制效率的同时增强了体数据特征的视觉感知.在绘制过程中,文中算法利用简单而有效的特征分析操作,在传输函数映射的基础上对感兴趣特征的视觉元素再次优化,实现不完全依赖于复杂传输函数设计的体数据特征增强可视化.最后通过大量实验,进一步验证了该算法的有效性与实用性.
Transfer function has proven as a key step for feature classification and volume visualization. With the increase of the dimension, the design of transfer function has proven as a time-consuming and complex task which is hard for users to conduct interactions. As the volume visualization method based on the ray analysis is able to help users achieve better visibility of volumetric features, it is still difficult to conduct real-time volume visualization since the GPU resources have been largely occupied. In this paper, we introduce a high-efficient volume visualization method for feature enhancement. The rate of scalar change is firstly employed to recognize the features of interest along viewing ray. The opacity values of unimportant features are decreased to enhance the visibility of features of interest. Then, we further take advantage of unsharp masking to enhance the visual perception of the features of interest, with the high-efficiency of the volume visualization maintained at the same time. In the proposed method, the ray analysis is simple and the visual properties can be further refined to enhance the visual perception of features of interest. The users are able to achieve better visibility of internal features of interest without completely relying on complex transfer functions. A large number ofcomparison results further demonstrate the effectiveness and application value of our high-efficient volume visualization method.
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2014年第10期1663-1672,共10页
Journal of Computer-Aided Design & Computer Graphics
基金
国家自然科学基金(61303133,61303134,61100084)
浙江省公益技术研究工业项目(2014C31057)
浙江大学CAD&CG国家重点实验室开放课题(A1417)
国家“八六三”高技术研究发展计划(2012AA12A404)
关键词
体数据
传输函数
视觉感知
反锐化掩模
volume data
transfer function
visual perception
unsharp masking
