MC Algorithm And Mesh Simplification Algorithm Based On GPU And Hadoop

The 3D reconstruction of medical image is a kind of 3D image reconstruction and visualization,which is based on visualization technology. It has important research and application value. MarchingCubes algorithm is widely used in 3D reconstruction,but it can generate large scale triangular mesh.Therefore,the ability of real-time rendering for mesh simplification can be improved without affecting the quality of the model.In this paper, we propose a method of moving cube based on vertex state for the problem of empty cube in MarchingCubes algorithm. According to the state of the vertex, we determine whether or not the adjacent body element contains the equivalent edge and determine the next cube with the isosurface, so as to avoid the traversal of the empty cube. The method is parallelized in GPU and Hadoop environment to improve the speed of 3D reconstruction. This paper focuses on the problem of model simplification based on QEM, which does not pay attention to the model region. Based on the QEM, this paper proposes a new edge folding algorithm based on region feature,which is based on the curvature and the length of the folded edge. The triangular mesh simplification based on region feature is parallelized,which the region features of the original model are reserved. The experimental results show that the MC algorithm based on vertex state achieves a faster reconstruction speed. In addition, the edge folding algorithm based on region feature preserves the important features of the model better and improves the real-time rendering speed. And the processing speed of the algorithms after parallelization is improved.