Interpolation of scattered data: investigating alternatives for the modified Shepard method

Many scientific applications use scattered data originating from samples. Interpolation techniques are necessary to estimate the values on non-sampled regions. In previous work, the modified Shepard interpolation method was implemented and some unpredictable results occurred. Due to the flexibility of parameter definition for interpolant generation in this method we decided to investigate solutions for those problems. In this paper we present four versions for the modified Shepard method, and compare their results through the analysis of the images, processing time and accuracy. Future work for further improvements of the method is also discussed.