A discrete 3D binary approach to compute the skeleton in 2D grey-level images is presented. The 2D grey-level input image is converted to a 3D binary image and the top surface of the foreground is identified. This discrete surface then undergoes skeletonization. The obtained 3D curve skeleton is pruned, before being projected back to a 2D grey-level image. This is suitably post-processed, since the projection may cause spurious loops and thickening. This algorithm can find applications in optical character recognition and document analysis or in other situations where shape analysis by skeletons is desired. An important property of the suggested method is that no hard segmentation into foreground and background is needed prior to the skeletonization.