scientific article; zbMATH DE number 1167539

From MaRDI portal
Publication:4396797





The problem of inversion of the Radon transform is treated via reconstruction methods, convergent in the uniform norm and based on the concept of reproducing kernels. The fact that the Radon transform can be ``normalized to map polynomials onto polynomials is used, so that one can get all the information by studying the transform's action on polynomials. The author shows that all the orthogonal projections, naturally occurring, can be identified with the orthogonal projection onto a certain subspace of the space of spherical harmonics with two additional dimensions. Thus some classical results on summation of Laplace series are applicable to the inversion problem for the Radon transform. Namely, row-finite matrix-transforms, making the Laplace series of a continuous function convergent in the uniform norm, are considered. Every such matrix-transform acts in the same way on the corresponding orthogonal projection series of the function to be reconstructed, thus defining a ``filter, by the use of which the reconstruction on the ball can be performed with arbitrary precision. The approximants are polynomial restrictions. A reconstruction problem of tomography is considered.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4396797)