The 1-D Hermite Shepard and MLS method

In many applications, one encounters the problem of approximating 1-D curve and 2-D surfaces from data given on a set of scattered points. Meshless methods strategy is based on some facts: (1) deleting mesh generation and re-meshing, (2) raising smooth degree of solution, (3) localization by using compact support weights. This research presented three generalizations for ancient pseudo interpolation, localization, appending a complete polynomial to the Shepard average weighted approximation and Hermite form of Shepard and MLS method. The new bases for relevant space of approximants are developed and, when evaluated directly, improves the accuracy of evaluation of the fitted method, especially the Hermite type. In this work, we develop some efficient schemes for computing global or local approximation curves and surfaces interpolating a given smooth data. Then we raise the smooth degree of approximation and use of derivatives data. The Hermite Shepard (HSH) is straightforward and efficient.