Medical Image Inpainting with RBF Interpolation Technique

Abstract

Inpainting is a method for repairing damaged images or to remove unwanted parts of an image. While this process has been performed by professional artists in the past, today, the use of this technology is emerging in the medical area – especially in the medical imaging realm. In this study, the proposed inpainting method uses a radial basis function (RBF) interpolation technique. We first explain radial basis functions and then, the RBF interpolation system. This technique generally depends on matrix processes. Thus, matrix operations are executed after the interpolation and form a main part of the process. This interpolation matrix has a known value used for interpolating n values and we need to find M – 1 for the n + 1 values of the original and inserted data. Implementation of an inpainting operation is carried out in an object-oriented programming (OOP) language. This is where the process is completed. The algorithm used in this study has a graphical user interface. Further, several skin images are used for testing this system. The obtained output represents a high level of accuracy that can support the validity of the proposed method.