Natural Neighborhood Algorithm for Kriging Interpolation on Medical Images

The study of images with spatial properties has been a trend in mitigating and comprehending information from the large and dense image data sets. The grabbed images need various types of corrections related to elevation, presentation and proper interpretation of the spatial features. Inpainting tools help in correcting and recovering the damaged portions of the images. Medical images as they are from complex medical imaging systems are always prone to external influences of light, heat and other hardware overheads. The inpainting is the mechanism that corrects and eliminates the damaged portions of the images using kriging and interpolation. The new method called natural neighborhood interpolation is introduced in order to reduce the burden of the variograms design of the traditional kriging process. In this paper, we introduced the natural neighbor interpolation and found that it is a better approach compared to the traditional kriging processes. The experiment has been tested on the synthetic medical images.