Enhancement of the Lohmann-type computer-generated hologram encoded by direct multilevel search algorithm

Abstract

An optimization-coding algorithm is required to produce high-quality computer-generated holograms (CGHs). For such a purpose, we have developed the direct multilevel search (DMS) algorithm using the rigorous reconstructed-image representation for encoding Lohmann-type CGHs. However, it required much time owing to the one-by-one search process: the optimal amplitude and phase must be selected from many multilevel amplitudes and phases. To solve this problem, we have established the AE1-trial search method, which can greatly reduce the trial search number: searches are limited only to the minimum shifts in amplitude and phase instead of all the possible shifts. Moreover, the DMS algorithm that incorporates the oversampled encoding method has been verified to be effective for reducing the optical reconstructed-image error by preparing the CGHs using electron-beam lithography.