Flow Visualization Based On The Model Of Perception

The flow visualization of scientific computing visualization is a classical research field, andis of great significance and value on marine scientific research and analysis. Plane vector fieldvisualization is an important part of visualization in scientific computing. At home and abroad,vector field visualization methods and techniques have been divided into the following types:direct visualization, geometry-based visualization, texture-based visualization, structure-basedvisualization. Texture-based flow visualization is a widely used method which shows thecharacteristics of the vector field by computing with a noise texture. Many domestic and foreignscholars and experts of the flow visualization have proposed many excellent algorithms such asthe line integral convolution method and the spot noise visualization which have been applied inmany areas.Scientists have always tried to contribute to the algorithms to produce more effective datavisualizations, so how to evaluate the quality of visualization is also a new visualization researchdirection. The ultimate goal of visualization is to enhance the user’s perception of the process ofvisualization, so the human factors are the important factors that can’t be ignored. With theresearch and development on the theory of biological, scientists have proposed a mathematicalmodel for simulating the primary visual system, which makes the evaluation of the quality ofvisualization and quantitative analysis become a reality.Based on the improved line integral convolution algorithm, this article proposed amathematical model of human biological visual perception system to simulate the process ofuser-aware from the results of visualization, and reconstruct an intermediate vector field of thesame resolution with the source vector field. In order to evaluate the accuracy of the visualization,we use the conditional entropy of the two vector fields as the similarity. So the parameters of thevisualization methods will be adjusted until the corresponding conditional entropy get itsminimum. In order to explain the correctness of the evaluation method, we design apsychophysical experiment with the user participation, combining the application of subjective human visual perception with mathematical models to prove the validity of the evaluation andoptimization method.