Color digital hologram compression based on matching pursuit

Abstract

With the recent widespread interest for head-mounted displays applied to virtual or augmented reality, holography has been considered as an appealing technique for a revolutionary and natural 3D visualization system. However, due to the tremendous amount of data required by holograms and to the very different properties of holographic data compared to common imagery, compression of digital holograms is a highly challenging topic for researchers. In this study, we introduce a novel approach, to the best of our knowledge, for color hologram compression based on matching pursuit using an overcomplete Gabor's dictionary. A detailed framework, together with a GPU implementation, from hologram decomposition to bitstream generation, is studied, and the results are discussed and compared to existing hologram compression algorithms.

JPEG Pleno: Providing representation interoperability for holographic applications and devices

“… and El Rhammad et al. using Gabor/Morlet wavelet dictionaries and a matching pursuit approach, providing a practical tool actually exhibiting the duality between light rays and local spatial frequencies.…”

“…Lee et al [40] used H.264/AVC to encode the numerical reconstructions of sub-holograms corresponding to specific viewing angles. Extending this approach, directional scalability was investigated by Viswanathan et al [41] and El Rhammad et al [42,43] using Gabor/Morlet wavelet dictionaries and a matching pursuit approach, providing a practical tool actually exhibiting the duality between light rays and local spatial frequencies.…”

Progressive hologram transmission using a view-dependent scalable compression scheme

“…More specified wavelets schemes that match the statistical properties of holographic signals were designed using Bandelets transform [17] and a vector lifting scheme [18,19]. Another category of works considered the use of B-splines/Fresnelet [20,21] and Gabor/Morlet [22,23] wavelets as an efficient tool to deal with non-local nature of holograms due to their optimal space-frequency localization property. Recently, a new compression paradigm was introduced in [24], using Linear Canonical Transforms to construct a piece-wise operator that models the diffraction of non-planar surfaces.…”

“…Unfortunately, a strong speckle noise is caused by these approaches due to the small aperture size. To overcome this limitation, a view-dependent compression scheme using Gabor wavelets to extract the light rays emitted in the viewer's pupil is introduced in [31]. Despite the good performance achieved by this technique in terms of compression and viewpoint scalability, the study did not take into account a progressive increase of quality for each sub-hologram.…”

Recent Advances in the Processing and Rendering Algorithms for Computer-Generated Holography

“…The transform coding with well-known transforms has also been investigated, using the DWT (Discrete Wavelet Transform) [16] and the DCT (Discrete Cosine Transform) [17,18] transforms. To increase the holographic content-awareness during the transform stage, other types of transforms have also been proposed, such as Fresnelets and Gabor wavelets [19,20]. Furthermore, variation of standard still-image and video codecs have been proposed and investigated to improve compression performance on holographic data.Some examples include the JPEG 2000 variation proposed in [21] and the HEVC extension proposed in [22].…”