Research And Application On Related Techniques Of 3D Reconstruction Based On CT Data

With the development of computerized tomography and nuclear magnetic resonance, three dimensional reconstruction based on layer data have become one of the most popular research topics, it is related to the research fields of medicine, digital image processing, computer graphics, computer vision, and so on. It is now widely used. This dissertation researches the problem of three dimensional reconstruction based on CT data, and apply it in facial reconstruction.It mainly includes:1) Contours extract and simplification .Adopt threshold methods in contours extract after image smooth, image tone; Propose a method of reliable lines-residue points, first, find the reliable lines of the contours , then combined this lines with residual points of the contour ,gained the characteristic points of the curve, using this points to fit the primary curve.2)Three dimension reconstruction based on CT data. For our reconstruction problem, a new method that align the particle of the contours in adjacent CT slices is presented to locate the contours, then the shortest diagonal method is used to get the model of the skull and skin.3)Triangle mesh simplification and display. Many triangles should be stored after the reconstruction, it affects the calculation and display of the model . The method of edge collapse based on quadric error metrics is improved, this can reduce the triangle that is long and narrow after the simplification, and it also make the storage tail off. Adopt graphics warehouse to display our model.4) A three dimensional reconstruction system based on CT data is built. Realizing the input and pretreatment of the data , contour extract and simplification, surface reconstruction, triangle collapse, model display and so on.