| At present, 3D reconstruction from medical images is a hot subject in the field of Visualization in Scientific Computing. It is an important application of computer graphics and image processing to biomedicine. 3D reconstruction will be applied to medicine, geology and aerography widely. 3D surface reconstruction based on medical images is a primary topic of visualization of 3D data sets. Reconstruction of 3D model is the premise of analysis and emulation. Based on mis background, the paper mainly studies 3D surface reconstruction from medical images and related techniques.The main contents of 3D surface reconstruction from medical images include image pre-processing, segmenting and extracting tissues or organs, constructing 3D surface models, and simplifying the mesh of the models.In this thesis, some key issues of 3D surface reconstruction from medical images are systematically and deeply investigated. Firstly, contrast reinforcement, filter and some other pre-processing techniques of medical image slices are reviewed. An image reinforcement method is proposed based on linear dynamic range adjustment which can improve the quality of images enormously. Then, on the bases of those, the study of contour extraction is carried out. A method for contour extraction, based on Log differential operator, is presented and the contour is condensed by the reservation of characteristic points. It is proved that the algorithm is efficient for extraction and condensation of contours. Finally, 3D surface reconstruction based on 2D contours is discussed mainly. The methods of reconstruction 3D objects from single branching 2D contours are presented and the problem of correspondence and branching in reconstruction 3D objects from multiple branching 2D contours are discussed in detail. A concrete scheme for those problems is presented. A plenty of experiments of 3D surface reconstruction from medical images illustrate the effectiveness of the algorithms presented in the paper.
|
PDF Full Text Request