Fast Computational Integral Imaging Reconstruction Method Using a Fractional Delay Filter

In this paper, we propose a fast method for computational integral imaging reconstruction (CIIR) in computational integral imaging (CInIm). The previous CIIR method requires a very high computational cost to reconstruct a volumetric image from elemental images. To overcome this problem, we introduce a fast technique that reduces the computational cost markedly in comparison with the previous method. The proposed method employs a fractional delay (FD) filter in place of the magnification of elemental images. Then, the FD-filtered elemental images are superimposed on each other. We calculate the numbers of pixels that are considered in our CIIR method and in the previous one. On the basis of this, the proposed method can be a fast process because the number of pixels from our method is much less than that of pixels from the previous one. To show the usefulness of the proposed method, we carry out preliminary experiments and present the results in terms of computational cost and visual quality. (C) 2010 The Japan Society of Applied Physics