Three methods of information assessment for optical data processing

The determination of the specific information in a message has proved of interest in coding (for instance in a hologram) as well as in data processing itself. Also, simple considerations about the definition of optical information have grown into several groups of processing methods. At first the concept of characteristic spatial frequency bands, which describe essential features of any object, leads to compression of its Fourier terms in order to increase the transfer density of a transmitting system and the storage capacity of the recording media. With the emphasis being put on the informative value of the reference in holography, geometrical compression is applied in particular to the coding of printed texts. The second and third groups of methods apparently deal with opposite approaches. Assuming a given message, one may select some parts of its spectrum for further processing or carry out average signals from the collections it contains. Both cases-sampling and averaging-permit the extraction of a group of data belonging to a set of discrete signals. The following developments are presented: image replicating at a determined energy level, step-by-step phase sampling grids, study of the displacement law for moving objects by random sampling of disturbed wavefronts and recording of incoherent holograms, on the one hand, and construction of average letters for comparison between manuscripts, average Fourier holograms, and power spectra, on the other hand. A last illustration is provided in connection with the appraisal of psychological tests.