Both image reconstruction and restoration transform the digital signal from the image-gathering device into a picture. However, whereas the reconstruction is intended to produce a continuous representation of this signal, the restoration is intended to produce a representation of the scene (i.e., to minimize the distortions caused by the perturbations in the image-gathering process). Traditionally, image-gathering devices are designed for the best possible image reconstruction even if they are subsequently used with image restoration, and image-restoration algorithms fail to account for the insufficient sampling (aliasing) in the image-gathering process. By informationally optimizing the design of the image-gathering device and properly accounting for its perturbations by the image-restoration algorithm, it is possible to obtain sharper and clearer pictures at as lower data rate. The cited information rate and the theoretical minimum data rates are for random scenes with mean spatial detail equal to the sampling intervals.