Design Transfer Function To Implement Context-Presserving Volume Rendering

Scientific visualization is the discipline of automatically rendering images from scientific data. Adequate visual abstractions are important to show relevant information in the data. Medical visualization have important meaning in the medical science study. The volume rendering techniques are the important point of Scientific visualization. The volume rendering techniques included direct volume rendering and indirect volume rendering. Indirect volume rendering generated an intermediate representation of the dataset that is rendered afterwards. Direct volume rendering provides a different strategy; here, the volumetric dataset is directly represented visually without generating a meta representation. Furthermore, the complexity of direct volume rendering algorithms is driven by the number of voxels of the dataset and by the number of pixels of the viewing plane. Direct volume rendering because of the advantages of imaging results have been widely used and research.Ray-casting algorithm as one of the direct volume rendering methods, which can meet the quality requirements of image reconstruction, are widely used in the medical field. The result of ray-casting algorithm couldn't simultaneously visualize interior and exterior structures while preserving clear shape cues. Volume clipping and optimization transfer function methods are proposed to resolve these issues. Volume clipping allows us to cut away selected parts of the volume, based on the position of voxels in the data set. Very often, clipping is the only way to uncover important, otherwise hidden, details of a data set. But, clipping can also remove important context information leading to confusing and partly misleading result images. Optimization transfer function method by optimizing the contrast between different organizations, enhance visual discrimination.The methods of optimizing transfer function are analyzed and concluded, model depended on data itself and user's needs are proposed focusing on the disadvantages of these methods. Two different context-preserving volume rendering approachs are proposed, and the display effect is enhanced efficiently. The proposed model is accelerated by using calculation figure processor to accelerate the interaction speed of the two models. The main work are detailed as follows:(1) A context-preserving volume rendering approach for accurate medical image visualization and analysis is proposed, which can enhance the display of region of interesting (ROI) and preserve the clear shape cues of background. First, a specific iso-surface along the viewing ray is located. For a sampling point, its distance to this iso-surface and its gradient are computed. Then, an attenuation function is constructed using the distance and the gradient of the sampling point. Finally, the attenuation function is integrated into opacity transfer function to control the rendering. The proposed method is assessed on the head, abdomen, and feet CT data. The experimental results show that the display of ROI is efficiently enhanced while the contours of tissue are preserved.(2) Transfer function based on Gaussian function is provided according to client's choice of interesting area. User's choose the interesting point first and construct opacity transfer function then. Distance from the sampling point to this interest point is taken as input, angle between the connection line from viewpoint to the center of interest and the line of sight is used to control the size of perspective region, and the tissue's context information effectively retained, thus to get to the purpose of displaying the region of interest.(3) With the development of General Purpose Graphics Processing Unit, to test the performance on Ray-casting algorithm of GPGPU and construct applicable software platform, CUDA platform of NVIDIA company are chosen to realize Ray-casting algorithm and the algorithm is encapsulated into an existing VTK software package for secondary development.