Research On Flow Field Visualization And Immersive Simulation Technology For Perception Enhancement

Flow visualization plays an underlying role in helping users to analyze and comprehend complex flow mechanism,provides insight into flow physical phenomena and discovers flow scientific laws.This thesis aimed at studying of the major difficulties in the flow visualization process,enhanced perception of flow physical process was regarded as the research target,theoretical innovation and technology breakthrough were earied out from visual rendering,rendering environment,interaction technique,layout and other aspects.Our method could extend the variety of information expression in the flow visualization,provide a favorable flow field visualization analysis environment to improve processing efficiency of flow scientific data and visual analysis ability.The main contributions of this thesis can be concluded as follows:1.The classification strategy for perception enhanced flow visualization based on VCIH model was proposed by drawing from the experience of domain experts analyzing flow physical phenomena,then the research about flow visualization methods were summarized based on perception enhanced from four aspects:enhancement of visual perception,enhancement of flow physical process,enhancement of exploratory interaction and enhancement of hardware dependencies.2.In order to convey the actual vector field structure carried by the data and to improve the rendering quality and visual effects of the streamline,a strategy for adaptive streamline based on geometrical characteristic in 3D flow field space was presented to show a clear flow field pattern and the important flow features in our algorithm.In order to have a nice expressive force in the visualization of stream surface,an adaptive triangulation technique was adopted to keep the structural continuity,smoothness,concavity and convexity of stream surface,and it could better reflect the underlying properties of the flow field and significantly improve the readability of 3D flow field dataset.3.In order to improve the quality and compatibility of rendering,a novel ColorLIC rendering technology based on hierarchical clustering combined with color feature mapping was proposed,the visual presentation of detailed features could be improved effectively by global sampling and neighbor clustering algorithm.In order to solve the problem of occlusion and confusion existed between texture slices caused by spatial characteristic when extended to 3D flow field,a new sparse noise based improved VolumeLIC rendering technology of 3D vector field implemented by GPU acceleration was proposed in this thesis.It could meet the requirements of the real-time display and gave an effective way to improve the structural sharpness and visual expression of 3D texture rendering.4.Combing with the analytical advantages and interactive characteristics under the immersive environment,the design principles of immersive interaction were defined.According to the principles,a novel non-contact gesture interaction method was proposed to apply in the immersive environment.The evaluation model of motion comfort was introduced to design the interactive hand gestures,the spring model was used to eliminate the phenomenon of hands tremble,and the cursor model was defined and applied to estimating the gesture state combined with context gestural motions.Finally,the test platform was established with HTC Vive+Leap Motion to verify the advantage of interactive experience of our method.5.Aimed at the problem that single visualization method cannot meet the cognitive need of analyzing and comprehending complex flow mechanism and flow physical phenomena,the multiple-view collaborative interaction model was introduced for parallel rendering of the statistical eharacteristics of flow attributes,eritical state distribution and spatial-temporal evolution laws from a multi-dimensional perspective.The multi-dimensional association and contrast model was introduced to render visual effects in different environmental conditions,and it could help to realize the all-round and multi-angle observation,comparison and analysis of multi-dimensional attribute data.A new visualization system for large scale CFD flow field data called FVIS was designed and implemented to verify the effectiveness of our approach.Use cases combined with expert opinions verified the fiinctionality and effectiveness of the proposed layouts,thus substantially increased the efficiency and accuracy of the analysis with complex multivariate flow data sets and the potential association relationships among them.