Algorithm For Extracting Skeleton Of3D Mesh Model Based On Reeb Graphic

In recent years，the study of mesh skeleton extraction becomes a rearch hot topic.There aremany thorough rearches on skeleton ectraction and its application done by both domestic andforeign institutions as well as scholars.These rearches have been gradully improved on theextraction results.Mesh model has been widely used in related areas such as3D animation, grid retrieval,medical perspective circuit directions because of its skeleton maintained a better characteristic intopology structure. In that way it has become the study focus that how to quickly extract skeletonaccurately. This paper summed up several skeleton ectraction methods and following proposedan improved triangle mesh model skeleton extraction algorithm on Reeb ideology. The model isfirst preprocessed to a simpler one premised the invariance of topology structure.Then thefeature point is calculated on the simplified model and geodesic distance function is added inclassfication function where the feature point conducts category coordinate point. The featurearea obtained by the function values of different vertices and using mesh vertex information ofdifferent branches to topology analysis.The vertices in same area grouped to a skeletonpoint,then togethered the adjacent skeleton point following the principle of proximity if there isnot only one skeleton point in same area,finally the skeleton point connected between theadjacent topology structure.This step prevented producing excess skeleton branches. The finalimplementation showed that the skeleton obtained by the improved algorithm can get a bettertopology structureThe research work mainly includes the following aspects:1.Several kinds of methods of triangle mesh skeleton extraction are studyed,at the same timethe advantages and disadvantages of each method are summarized.Thus we determine the topicthat Reeb grapg-based skeleton extraction algorithm to extract skeleton of the model. It isproposed that whole idea of the algorithm, scheme and solutions to the main problem.2. Previous skeleton extraction algorithms operate on grid model directly, which ignore thesituation that has a large number of vertices or overly complicated. In order to reducecomputational complexity,the model is first preprocessed to a simpler one and then is operatedby the characteristic-keeping mesh simplification algorithm Thus the data of grid vertex and faceare reduced and lead to a lower time and space complexity for subsequent calculation withoutchange the topological shape.3. Skeleton extraction algorithms based on Reeb Figure are classified by the same discretecontour vertices. This paper puts forward a new method based on Dijkstra Algorithm to calculatethe geodesic distance of grid vertex. More precise results can be obtained by the new method.When calculating the characteristic points,we can get the characteristic point set by firstcalculating the set of two different characteristic points and then calculating their intersection.Adetailed implementation is proposed in this paper.4. The final skeleton point is extracted by the method of vertex polymerization, whichselected a geodesic distance function classifying the mesh vertex according to the principle thatvertices belonging to the same classification function value are classified as a class.By topologyanalysis of the vertex, the skeleton point obtained for each area. And finally the topologyconnection of the topology adjacent skeleton can get the modle skeleton line. It can be shown that this algorithm can get a more precise skeleton to reflect the character of model topology byfinal implementation.