Performance assessment of a fast and accurate scalar optical diffraction field computation algorithm

Abstract

Dynamic holographic reconstructions can be obtained by employing digital holographic video displays which are pixelated devices. In practice, spatial light modulators (SLMs) are used in such purposes. The pixelated structure of SLMs can affect quality of the reconstructed objects. Hence, to obtain better reconstructions, pixelated structure of SLMs has to be taken into consideration. Rapid calculation of the diffraction field which is emitted by the object is just as important as the accuracy of the diffraction field. The presented algorithm is based on computation of Fresnel integral over each pixel area on SLM, thus accurate results are attained. Fast computation of the diffraction field is obtained by scaling of a pre-computed diffraction field compared to standard way of computing the diffraction field. Scaling operation is obtained by using three interpolation methods: rounding to nearest value, linear and cubic interpolation. Although, the rounding to nearest value gives the shortest computation time among all three interpolation methods, it provides the largest normalized mean square error (NMSE). The smallest NMSE can be attained when cubic interpolation is used for computation of diffraction field, but it yields the longest the computation time. Even if NMSE performance of the linear interpolation method is not as good as the cubic interpolation method, the computation time can be reduced by nearly half. [Figure not available: see fulltext.] © 2013 3D Display Research Center and Springer-Verlag Berlin Heidelberg.