In this paper we consider landmark-based image registration using radial basis function interpolation schemes. More precisely, we analyze some landmark-based image transformations defined by means of compactly supported radial basis functions, namely multivariate Wendland’s and Gneiting’s functions. The latter, as far as we know, have never been used in this context. Comparisons of the two transformations are given. Numerical experiments performed on test examples show better accuracy of Gneiting’s ones in some cases. Finally, an application of compactly supported transformations to real medical images is also considered.