|
Three-dimensional reconstruction of medical images is an important research direction in medical visualization. The three-dimensional reconstruction by two-dimensional tomographic images to reconstruct the three-dimensional entity, to provide to the user to observe and interact. Through three-dimensional reconstruction of medical images, can accurately reflect the spatial location of human tissues, organs, size, intuitive and complete information to the doctor, to help improve the efficiency and accuracy of the diagnosis. The in-depth analysis the MC surface rendering algorithm based on ray casting volume rendering algorithm improved rendering algorithm, based on the characteristics of medical CT images, the use of the VTK visualization tools to achieve a three-dimensional reconstruction based on CT images. Of this study and the analysis of the VTK visualization tools, the characteristics of the VTK visualization tools, constitute, and class library structure. The paper discusses the CT image preprocessing technology for medical image acquisition process may be caused by noise, the use of neighborhood mean filter and median filtering of image smoothing, Laplacian sharpening of image enhancement processing. Adjacent to the CT image is much larger than the distance between the two layers of a tomographic image with a tomographic image of the adjacent distance between the two pixels, which makes direct reconstruction of three-dimensional entity to a greater difference with the real object, the paper interlayer, according to a relatively large 3D medical data sets using the linear weighted average interpolation. Roberts, Prewitt, Sobel three edge split operator on the CT image segmentation, extract the object of interest, and to prepare for the three-dimensional reconstruction. Marching Cubes surface rendering algorithm analysis, the traditional the MC algorithm needs to traverse all voxels in the field of three-dimensional data, algorithms, most of the time spent on the detection of empty voxels affected the efficiency of the implementation of the algorithm. In this paper, to solve this problem isosurface search method based on seed filling algorithm to reduce the access to, and improve the algorithm execution speed of empty voxels. Rendering algorithm based on the body of light projected big problem for the need to change the viewing angle display re-sampling new perspective and lighting conditions, the amount of computation, this paper gives an improved body rendering algorithm: Use surrounded box techniques for cutting the three-dimensional data field, to reduce the amount of data to be accessed, while the two-dimensional Bresenham algorithm is extended to three-dimensional, and the light emitted in the volume rendering discrete voxel, and using the front to back of the image compositing The algorithm colors synthesis, ahead of the end of the light, and to improve the reconstruction speed. According to the experimental results opposite these two categories of rendering and volume rendering three-dimensional reconstruction algorithm were compared and analyzed, discussed their areas of application. Finally this paper, the use of the VTK completion of the three-dimensional reconstruction system, the main functions were introduced.
|