Research On High Efficiency Ray Casting Algorithm Based On Data Classification

Scientific computing visualization converts 3D data into images with rich information and interacts with them.Visualization of underwater acoustic data as one important branch of that,mainly used in underwater resource detection,shipwreck salvage and fishery applications.Due to the complexity of the underwater environment and the data collected by the sonar are susceptible to noise,effective mapping is the core work of underwater acoustic data visualization.Volume rendering technology occupies an important position in the development of 3D underwater acoustic data visualization.Therefore,after introducing several volume rendering algorithms,the most typical ray projection algorithm is selected to complete the rendering,especially for the design of the most important transfer function,and the visualization platform has been further designed and developed.The main research contents and results are as follows:The key to the process of visualizing is the design of the transfer function,which aims to highlight important information and structural features.The underwater acoustic data is mainly composed of the water layer,the transition layer and the stratum.Since the target data is included,the boundary information is also crucial.The LH histogram transfer function can significantly improve the distinguishing ability of the boundary.However,the original LH value calculation range is large and cannot meet the requirements of high efficiency.Therefore,a design method of LH histogram transfer function based on region space is proposed.By selecting a certain region space in the underwater acoustic data,and calculating the mean and variance of each voxel in this region and its neighboring voxels,by comparing the variance and set the estimated range value to determine whether it is similar to its neighbor voxels to determine whether it is an internal equivalent voxel,thus reducing the computational workload of FL and FH values,and then traversing all the spatial space of the volume data to obtain the regional space LH histogram Graph transfer functions to achieve the need for efficient classification mapping.The experimental results show that compared with the existing threshold-based LH histogram transfer function method,the spatial space-based LH histogram method saves nearly 50% of the computational cost,while maintaining good boundary distinguishing ability and better rendering result,effectively improving the problem of long calculation time of the light projection algorithm.Combined with the cross-platform C++ graphical user interface library Qt and visualization toolkit VTK,designed and implemented underwater acoustic data application platform based on VTK and Qt.It uses VTK for image visualization,and then uses QT to design and interact withgraphical user interface to design two classics scalar value transfer function and the gradient value transfer function control,and the underwater environment is truly restored by setting small controls such as interpolation,shadow,and light intensity.The experiment uses two different ray projection algorithms to test the two sets of sliced data after classification,by comparing the controls under different settings,so as to quickly draw a clear three-dimensional underwater acoustic image to achieve the interaction requirements.