vtk accumulation

1. Discrete isolated point processing method. Note that there are many small,
   disconnected isosurfaces due to noise and moisture isolated in the data. We use vtkConnectivityFilter to remove these small, disconnected surfaces. Figure 9.51 b
   shows the result of applying the filter. More than 50,000 triangles were deleted, leaving only 299,480 triangles.
   vtkConnectivityFilter is a general-purpose filter that takes a dataset as input and generates an unstructured grid as output.
   Its function is to extract cells connected at a point (that is, share a common point). In this example, single
   extracts the largest surface. You can also specify cell id and point id, and extract the surface connected to them.
   Laplace's smoothing pipeline, vtkSmoothPolyDataFilter::SetNumberOfIterations() controls the number of smoothing. The larger the number, the more smoothing is, that is, the more detail loss. There are also many variables in this category to control the smoothing process, and the loss of detail can be controlled to a certain extent by using these changes. BoundarySmoothing controls whether to smooth the boundary points. FeatureEdgeSmoothing controls whether to smooth the points on the edge of the feature. If an edge is shared by two adjacent polygons, if the included angle (characteristic angle) of the normal vectors of the two polygons is greater than the defined threshold, then this side is a characteristic edge. Although the feature edge smoothing setting can reduce a part of the loss of detail, it cannot be completely avoided. With the continuous iteration of Laplacian smoothing, the model will gradually shrink toward the center of the mesh.
   And vtkWindowedSincPolyData is a better choice. The algorithm uses the window Sinc function to achieve mesh smoothing, which can minimize shrinkage.