The Research And Implementation Of Multiple Branching2D Contours On A Single Human Slice Image

With the development of science and technology, and people’s in-depth research onvisualization technology, the market has showed growing demands for virtual human, thethree-dimensional demic medoling technology would be widely applied in varieties of fieldsthat related to our human beings. We chould get more accurate data with the development ofcomputer hardware and software technologies and also the information acquisition andprocessing methods. These highly-accurate data would provide a solid and reliablefoundation on three-dimensional reconstruction of the human body and simulate the behaviorof a natural person in the future.Demic slice images that filled of demic visual data sets were widely used in medicalfield to achieve three-dimensional reconstruction. This paper focused on setting visualize dataas the surface modeling data material about human body. Firstly,we must get the data pointsof the body surface. The slice figures we get commonly contains a lot of redundantinformations due to the special environment of the the medical field, so slicing graphicsprocessing step was very essential. Images showed clear demic edge profile can be got afterpretreatment mainly for filtering noise, enhancement, segmentation and feature extractionoperations.Finally, this paper introduced a three-dimensional body contour modeling method,including multi-contour extraction algorithm and curve fitting algorithms. In the multi-contour extraction process, firstly extract the boundarycontour with a boundary trackingalgorithm and then determine the relationship between the contour line and point through point set classification algorithm. And then chose a new feature point retain algorithms to extract thefeature points of demic profile contour. At last,this article introduced anew data filtration method that based on string value method that simplify the boundary contour lines. It has introduced the Hermite interpolation curve and contour curve modeling with it in the method of curve-fitting algorithm. Finally, we successfull y got the contour lines of demic chest slice and realized the multi-contours planar reconstruction in single image through programmed experiments.