基于多频稀疏噪声的流场运动方向可视化算法

Visualization of flow field motion direction based on multi-frequency sparse noise

针对现有流场可视化算法存在的矢量场运动指向二义性以及缺乏对矢量场速率区分等问题进行了研究,提出了一种基于多频稀疏噪声纹理的改进线积分卷积算法,用来提高流场可视化后的信息表达效果。该方法首先根据矢量场速率量级的大小划分频率不同的噪声群,并合成适用于该向量场的特定多频噪声纹理;然后使用斜坡卷积核根据向量场流线对噪声纹理进行卷积;最后生成一幅同时具有矢量场运动方向和运动速率信息的彩色图像。实验结果表明,该方法可以增强用户对流场流量大小和矢量场方向的感知效果,在大规模风场数据可视化的验证实验上取得了良好的可视化效果。

Aiming at the problems that the flow field visualization algorithm existed the referential ambiguity of vector movement and lack of the distinction between vector field rate,this paper proposed an improved line integral convolution algorithm based on multi-frequency sparse noise texture to improve the information expression effect of flow field visualization.Firstly,the method divided the noise groups with different frequencies according to the magnitude of the vector field rate order,and synthesized the specific multi-frequency noise texture suitable for the vector field.Then,it used a slope convolution kernel to convolve the noise textures from vector field streamlines.Finally,it generated a color image with both direction and rate information of vector field motion.The experimental results show that this method can enhance the user’s perception of flow field size and vector field direction,and has achieved good visualization results in large-scale wind field data visualization verification experiments.