Research And Application Of The Complex Flow Field Visualization Based On Features

Flow visualization methods can turn abstract data into an intuitive image, which is an important branch of visualization in computer science. The line integral convolution algorithm (LIC), as a classic visualization method based on texture, can reveal the distribution of flow globally, and cannot lose the important information of flow field, which has extensive value in research and application.For the time-consuming problem of original LIC, a streamline enhanced LIC algorithm is put forward to make full use of the features’distribution of flow field. First, the algorithm should detect the position of features, and generate a gradient area. The important improved part is to generate the streamline of LIC by combining with different integration methods based on the region of feature points. Within the gradient region, the integral method of streamline should be more elaborate, because the flow changes drastically. Namely, an accurate method, the forth order of Runge-Kutta approach, is used within the gradient region, and Euler method with a constant integral step is used out of the area. Besides, the integral step is undated dynamically according to the position of current integral point within the area of the feature points. The nearer to the distance feature point, the shorter the step becomes. The integral step increase linearly from the feature points to the edge. In order to reduce the time-consuming, put the part of streamline on the GPU fragment processors. Finally, the image enhancement algorithm is used to increase the sharpness of visualization result further. This part is also run on the GPU, thus having less influence to the overall computing performance.In order to illustrate the adaptability of the improved algorithm further, this algorithm is applied onto the river flow data. A river flow field visualization system is constructed to display the result of flow conveniently and interactively. This system mainly includes the following functions. The bump texture is used to simulate river flow dynamically. And LIC algorithm is used to show flow distribution of the river surface statically. Besides, the system increases the terrain roaming function to observer environment surrounding by rivers.