Based On Pc-ct Image 3d Reconstruction Algorithm And Realization Of Research

At present, 3D reconstruction from medical images is a hot subject in the field of Visualization in Scientific Computing, and it will be applied widely in medicine. Based on a general research on methods and survey of 3D reconstruction from medical images, the dissertation studies mainly registration, filtering, interpolation and volume rendering technology from CT medical images. Many algorithms are realized in VC6.0 and OpenGL, which is based on hardware speeding up. A 3D reconstruction advance algorithm from medical images is realized, and 3D reconstruction system is developed for person computer instead of big workstation. The dissertation makes deep researches on the registration from medical images and the images automatic registration algorithm based matching feature points is presented. The computer experiments show that the algorithm is a fast, simple, efficient and accurate registration and the results are satisfactory. Then the dissertation does advanced research for medical images interpolation and corresponding point matching interpolation algorithm is presented, which is based on object feature. Compared with the linear interpolation, The new algorithm improves visual effect and its interpolation error. Based on the volume rendering algorithms, the dissertation presents Projection algorithm based on Voxel, in which the opacity parameter is introduced to achieve an 3D image with translucent effect. The dissertation synthesizes and absorbs the new research achievement of registration theory, interpolation theory and volume rendering technology. The dissertation realizes 3D reconstruction from CT medical images based on the person computer, and it will be applied widely in the future.