一种面向矢量场的纹理平流方法

A Texture Advection Method for Vector Field

为了避免利用稠密纹理可视化三维矢量场时造成内部结构遮挡及外部流模式的隐藏问题,文章提出一种基于噪声自适应采样的纹理平流方法。通过分析在三维矢量场局部区域运动模式上的相似性,自适应调整局部区域大小与点分布,在不同的视点条件下,充分考虑噪声样本点的距离及所在块对最终输出纹理的贡献调整噪声纹理的频谱,最终卷积合成平滑的流纹理轨迹。

In order to avoid the problem of internal structure occlusion and external flow pattern hiding when visualizing three-dimensional vector fields with dense textures, we develop a stream texture animation synthesis method based on adaptive sampling of noise. By analyzing the similarity in the local area motion pattern of the three-dimensional vector field, a heuristic algorithm is proposed to automatically adjust the size and distribution of the local area. Under different conditions of view, the distance of the noise sample point and the block where the output texture is located are fully considered. The contribution adjusts the spectrum of the noise texture and eventually convolves the smooth stream texture animation.