Correction des effets de l''image jumelle et du terme quadratique en holographie de Gabor

An iterative method to suitably solve the problem of the twin image and quadratic term in Gabor's holography is presented. This method gives useful information regarding the uniqueness of the solution and the sensitivity of the results to all types of errors. In particular, it appears that the hologram can be obtained in optical conditions intermediate between the Fraunhofer and the Fresnel conditions. This is of special interest in the case of Zernike-Gabor objects because of the signal-to-noise ratio. This technique seems to be well suited to the problems involved in electron microscopy. The applications concern the analysis at high resolution of structures in biology and in physics of condensed matter.