Implicit fitting of point cloud data using radial hermite basis functions

Normalized Implicit Radial Models for Scattered Point Cloud Data without Normal Vectors

“…Some related methods for knot selection have previously been used and discussed in [10], [13], [23], [25]. This method of selecting knots used here is based upon minimizing the distance between the knots { }…”

“…. Computational algorithms for the case of two-dimensions are discussed in [10] and for case of threedimensions in [25]. These algorithms involve the basic "venetia iteration" where a knot…”

“…which is the centroid of the data points lying in the Theissen/Dirichlet region of ( ) K j Q . As it is pointed out in both [10] and [25], this type of iteration quickly converges, but not necessarily always to a global optimal knot set that is a minimal distance to the data points. Nevertheless, we have found that with good initial knot distribution, this type of iteration leads to a very efficient means of determining knot locations which subsequently leads to overall efficient implicit approximation models.…”

DEM Void Filling Based on Context Attention Generation Model

“…However, the results of the above two methods depend on the sparsity and uniformity of the sampling points, while the algorithm is highly disturbed by known data. A radial basis function was originally used to interpolate scattered data [11]. The function and its improved form [12][13][14][15] were proposed, which had higher degrees of precision than interpolation fitting.…”