Based On The Volume Rendering Of Medical Image 3d Visualization Technology

The method of volume rendering can generate images that reflect the internal details of objects, and is becoming the mainstream of 3D visualization algorithm. However, as the large amount of computation, it is difficult to meet the requirement of real-time rendering. To address this issue, the shear-warp algorithm reduces the most time-consuming re-sampling process from 3D to 2D, and becomes the fastest volume rendering algorithm, having broad application prospects. But the algorithm still has some problems, such as the relatively large re-sampling step resulted in decreased image quality, and the volume data must be isotropic, but the isotropic processing may lead to loss of data information.In this thesis, the shear-warp algorithm is primarily studied and corresponding solutions for the problems within this algorithm are proposed. Experiments have verified the effectiveness of the solutions. The primary contents of this thesis are as follows:1. Researched and summarized the main volume rendering algorithms, especially the shear-warp algorithm. The most important concepts in this algorithm such as view matrix decomposition, coordinate system definition and conversion are further studied. Introduced a virtual sphere model which maps the 2D screen coordinate into a 3D space coordinate, and then the space rotation matrix can be computed.2. For the problem of declining image quality in shear-warp algorithm, proposed an improved method of image composition. In this method, the value of the light segment for compositing is decided by the two ends sampling points, not only the first point, and the value is more reasonable. Through simulation in Matlab, the method's effect on image quality improvement is verified.3. Presented a shear-warp algorithm based on the anisotropic volume data. For the data that is anisotropic but the spacing of the adjacent points is not great different in three directions, this method can render directly and avoid the information loss during the isotropic processing. Of course, this method is also applicable to isotropic volume data. 4. Completed the software implementation of the proposed algorithms, using a variable re-sampling rate, by which the requirements of rendering speed and image quality can be met at the same time.