An algorithm is proposed to image ionospheric electron density (IED) distribution. In this method, generalized singular value decomposition (GSVD) is first used to resolve the ill-conditioned problem in the computerized ionospheric tomography system. Its estimate is then provided as the initial approximation required by the improved algebraic reconstruction technique (IART). Numerical simulation has demonstrated that the combined algorithm is superior to both GSVD and IART for tomographic inversion of IED. Finally, the method is applied to perform inversion of IED using a set of global navigation satellite system (GNSS) data during a magnetically disturbed period. The reconstructed results reveal two prominent features of the ionosphere under the disturbed condition. The reliability of the method is also validated by the ionosonde data recorded at Wuhan station, China.