The Study On Medical Images Segmentation And 3D Reconstruction

As one of the hotspots in the field of computer vision, 3D reconstruction from medical images is a multi-disciplinary subject. It is an important application of computer graphics and image processing in biomedicine engineering. It relates to the subjects of digital image processing, computer graphics and some related knowledge of medical. 3D reconstruction and visualization of medical images are widely used in diagnostic, surgery planning and simulating, teaching in anatomy and medical simulate training. So the research on this issue is of high academic and applied value.The main contents of 3D surface reconstruction from medical images include image pre-processing, such as interpolating and filtering, segmenting and extracting tissues or organs, constructing 3D surface models, and simplifying of the mesh of the models.In this paper, we study some key techniques for tetrahedral 3D reconstructing from medical images (mainly for color anatomy images), we concentrate on medical color image segmentation and establish a set of methods to segment the medical images. The segmentation process consist of following steps: segmenting the medical images roughly, filtering the image using vector median filter, segmenting the medical images secondly by using of region merging algorithm, extracting contour of interesting region through region growing, extracting characteristic points and outputting them as a file of CNT format. Towards the filtering techniques, we study vector median filter for color image filtering. Based on the revised vector median filter, we present a new quick algorithm for color image filtering. As to the segmenting methods, we present two greedy clustering algorithms and a region merging algorithm which integrates information of color and space. The experiments for image processing algorithms we present in this paper show good performance.Moreover, we study the problem of 3d reconstruction of tetrahedral meshes in software Nuages. We also study and realize mesh smoothing for the reconstructed 3D model. Finally, a solution for multi-resolution presentation is proposed.