BINARY IMAGE RECONSTRUCTION FROM FOUR PROJECTIONS.

The problem of reconstructing an image from two pairs of its orthogonal projections is discussed. The range and support information of the image function is utilized in conjunction with the available projections to develop an iterative reconstruction method; the iterative algorithm is an error-energy-reduction scheme. Reconstruction results for two classes of image functions are presented. The scheme does not require alternate corrections in the spatial and spatial frequency domains of the image. As a result, the method is not as susceptible to noise and numerical errors compounded in the iteration process as are the currently available iterative reconstruction algorithms.