Research On Enhanced Vector Field Visualization Based On Line Integral Convolution

Vector field visualization technology mainly display the large-scale flow field data sets of scientific computing or simulation through the visual graphics and provide some means of human-computer interaction,So that researchers can intuitively and efficiently analyze the complex changes in the flow field,which is included in the data sets.In order to reflect the internal motion characteristics of the entire vector field,the texture visualization method based on line integral convolution is usually adopted,but the visualization result obtained by this method has low image quality.To solve this problem,this thesis optimizes the two aspects of texture enhancement and color enhancement,and designs an enhanced vector field visualization model based on line integral convolution.The main work of this thesis is as follows:1)According to the existing texture enhancement algorithm,the resulting image has obvious texture aliasing.Through theoretical research and experimental analysis on the key parameters that affect the result image quality,It is found that the texture aliasing phenomenon in the resulting image is caused by high-pass filtering using the equidistant sampling method,And this aliasing phenomenon has a relationship with the vector angle.Therefore,by analyzing the relationship between vector angle,sampling distance and texture aliasing,this thesis proposes a texture enhancement algorithm based on vector angle to adjust the sampling distance of high-pass filter,which greatly reduces the texture aliasing in the visualized image.2)The result image obtained by using the above texture enhancement algorithm can only show the directional information of the flow field and can not express the vector size.Usually,a linear color enhancement algorithm is used to add vector size information to a vector field,but the resulting image of the algorithm has a problem of color concentration.In view of this problem,this thesis analyzes the distribution characteristics of vector field using histogram,and proposes a dynamic non-linear color enhancement algorithm,which obviously improves the color distribution of result image and improves the visual quality of the result image.3)Under the same experimental environment,the existing algorithm and the algorithm in this thesis are compared and analyzed visually.Experiments show,the vector field visualization results obtained using this algorithm have high image quality and are more conducive to observe the details of the vector field.Therefore,this algorithm has a great practical value for the field visualization of various fields.