Downloads

HOLOGRAPHIC DATABASE

Here, you can find our public datasets and other resources from our research results.

CGH Dataset: Interfere-I

This data set accompanies our QoMex 2015 conference paper. If you report results of this benchmark, we request that you cite our paper entitled "Open Access Database for Experimental Validations of Holographic Compression Engines" for the 7th International Workshop on Quality of Multimedia Experience (QoMex 2015).
The dataset consists of 5 specular monochromatic holograms at a distance greater than 0.5 meters of 3D and flat objects. The CGHs of flat objects are computed with angular spectrum propagation, while the CGHs of 3D objects with an early version of our multiple WRP CGH method that integrates an occlusion processing technique.

NameNumerical reconstructionDescription Link
2D_Dice Hologram of 2D image of a Dice.
  • Resolution: 1920x1080
  • Pixel pitch: 8um
  • Wavelength: 632.8 nm
  • Reconstruction distance: 0.90 m
Download!
3D_Venus Hologram of 3D Point Cloud of a statue of Venus.
  • Resolution: 1920x1080
  • Pixel pitch: 8um
  • Wavelength: 632.8 nm
  • Reconstruction distance: 0.50 m
Point-cloud source used at courtesy of Direct Dimensions Inc. (www.dirdim.com)
Download!
2D_Multi Hologram of 2D Image of same dice in 3 different depths.
The images are taken from the point cloud of a 3D_dice.
  • Resolution: 1920x1080
  • Pixel pitch: 8um
  • Wavelength: 632.8 nm
  • Reconstruction distance: 0.51, 0.50, 0.49 m respectively
Download!
3D_Multi Hologram of 3D point cloud of same dice in 3 different depths.
  • Resolution: 1920x1080
  • Pixel pitch: 8um
  • Wavelength: 632.8 nm
  • Reconstruction distance: 0.51, 0.50, 0.49 m respectively
Download!
3D_Cat Hologram of 3D Point Cloud of a cat.
  • Resolution: 8192x8192
  • Pixel pitch: 2um
  • Wavelength: 632.8 nm
  • Reconstruction distance: 0.10 m
Point-cloud source used at courtesy of Tosca hi-resolution 3D database v1.0
Download!

Release: May 2015

CGH Dataset: Interfere-II

This data set accompanies our conference paper for SPIE Photonics Europe 2015. If you report results of this benchmark, we request that you cite our paper: Symeonidou, A., Blinder, D., Ahar, A., Schretter, C., Munteanu, A., and Schelkens, P., "Speckle noise reduction for computer generated holograms of objects with diffuse surfaces," Proc. SPIE 9896 (2016).

The dataset consists of 6 diffuse and 6 specular holograms generated from 3D point clouds with the multiple WRP method developed by our group. As described analytically in the accompanying paper, the diffuse CGHs were generated by assigning a random phase intensity per point, while the specular ones have only real intensity per point, equal to 1. The holograms are named with a D (diffuse) or S (specular), to denote the generation method. The holographic setup supports full parallax with a field of view of 37o. For the diffuse holograms the images shown are the numerical reconstruction of the right aperture (size: 2048x2048 pixels) of each hologram. The importance of random phase is proven by the smaller visible area in the specular version of each CGH.

For more information on the structure of the data, please read the Read_me.pdf file!

Diffuse holograms

NameNumerical reconstruction* DescriptionLink
Cat8KD Hologram of 3D Point Cloud of a cat.
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0140 - 0.0207 m
  • Reconstruction distance: 0.0142 m
Point-cloud source used at courtesy of Tosca hi-resolution 3D database v1.0
Download!
Size: 480 MB
Ball8KD Hologram of 3D Point Cloud of a perforated ball.
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0124 - 0.0176 m
  • Reconstruction distance: 0.0125 m
Point-cloud source used at courtesy of Manuel Piñeiro
Download!
Size: 480 MB
Venus8KD Hologram of 3D Point Cloud of a head sculpture of Venus.
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0128 - 0.0171 m
  • Reconstruction distance: 0.0129 m
Point-cloud source used at courtesy of Direct Dimensions Inc. (www.dirdim.com)
Download!
Size: 480 MB
Earth8KD Hologram of 3D Point Cloud of Earth.
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0118 - 0.0164 m
  • Reconstruction distance: 0.0118 m
Point-cloud source used at courtesy of Joerg Schmit
Download!
Size: 480 MB
Chess8KD Hologram of 3D point cloud of 7 chess pieces in 2 different depths.
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0135 - 0.0166 m
  • Reconstruction distance: 0.014 and 0.016 m front and back pieces respectively
Point-cloud source used at courtesy of Grace Chang
Download!
Size: 480 MB
Dragon8KD Hologram of 3D point cloud of a dragon
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0119 - 0.0180 m
  • Reconstruction distance: 0.0120 m
Point-cloud source used at courtesy of the Stanford 3D Scanning Repository
Download!
Size: 480 MB

* The images shown are the numerical reconstruction of the right aperture (size:2048x2048 pixels) of each hologram.

Specular holograms

NameNumerical reconstruction* DescriptionLink
Cat8KS Hologram of 3D Point Cloud of a cat.
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0140 - 0.0207 m
  • Reconstruction distance: 0.0142 m
Point-cloud source used at courtesy of Tosca hi-resolution 3D database v1.0
Download!
Size: 480 MB
Ball8KS Hologram of 3D Point Cloud of a perforated ball.
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0124 - 0.0176 m
  • Reconstruction distance: 0.0125 m
Point-cloud source used at courtesy of Manuel Piñeiro
Download!
Size: 480 MB
Venus8KS Hologram of 3D Point Cloud of a head sculpture of Venus.
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0128 - 0.0171 m
  • Reconstruction distance: 0.0129 m
Point-cloud source used at courtesy of Direct Dimensions Inc. (www.dirdim.com)
Download!
Size: 480 MB
Earth8KS Hologram of 3D Point Cloud of Earth.
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0118 - 0.0164 m
  • Reconstruction distance: 0.0118 m
Point-cloud source used at courtesy of Joerg Schmit
Download!
Size: 480 MB
Chess8KS Hologram of 3D point cloud of 7 chess pieces in 2 different depths.
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0135 - 0.0166 m
  • Reconstruction distance: 0.014 and 0.016 m front and back pieces respectively
Point-cloud source used at courtesy of Grace Chang
Download!
Size: 480 MB
Dragon8KS Hologram of 3D point cloud of a dragon
  • Resolution:8192x8192
  • Pixel pitch: 1um
  • Wavelength: 633 nm
  • Depth range of object: 0.0119 - 0.0180 m
  • Reconstruction distance: 0.0120 m
Point-cloud source used at courtesy of the Stanford 3D Scanning Repository
Download!
Size: 480 MB

Release: March 2016

CGH Dataset: Interfere-III

Colour holograms with Phong illumination model

This dataset consists of 6 colour CGHs. The set of 4 Sphere1A/B/C/D holograms show the same scene with different source of illumination, regarding colour and direction. “Sphere3” was generated by simulation of 3 spheres of different material as captured scene. The CGH Biplane is a hologram generated to demonstrate the possibilities of our CGH algorithm.

This data set accompanies our journal paper, published at Optics Express. If you report results of this benchmark, we request that you cite it as: A. Symeonidou, D. Blinder, P. Schelkens, "Colour computer-generated holography for point clouds utilizing the Phong illumination model," Opt.Express (2018)

For more information on the structure of the data, please read the Read_me.pdf file!

NameNumerical reconstruction DescriptionLink
Sphere1A Hologram of 3D Point Cloud of a red sphere with white light from South.
  • Resolution: 2048x2048
  • Pixel pitch: 8um
  • Wavelength: 633, 532 and 460 nm
  • Depth range of object: 0.295 - 0.305 m
  • Reconstruction distance: 0.2952 m
  • Point cloud: Download!.
Download!
Size: 90 MB
Sphere1B Hologram of 3D Point Cloud of a red sphere with green light from South-West.
  • Resolution: 2048x2048
  • Pixel pitch: 8um
  • Wavelength: 633, 532 and 460 nm
  • Depth range of object: 0.295 - 0.305 m
  • Reconstruction distance: 0.2952 m
  • Point cloud: Same as Sphere1A.
Download!
Size: 90 MB
Sphere1C Hologram of 3D Point Cloud of a red sphere with yellow light from South-East.
  • Resolution: 2048x2048
  • Pixel pitch: 8um
  • Wavelength: 633, 532 and 460 nm
  • Depth range of object: 0.295 - 0.305 m
  • Reconstruction distance: 0.2952 m
  • Point cloud: Same as Sphere1A
Download!
Size: 90 MB
Sphere1D Hologram of 3D Point Cloud of a red sphere with white light from South and blue from North-West.
  • Resolution: 2048x2048
  • Pixel pitch: 8um
  • Wavelength: 633, 532 and 460 nm
  • Depth range of object: 0.295 - 0.305 m
  • Reconstruction distance: 0.2952 m
    Point cloud: Same as Sphere1A
Download!
Size: 90 MB
Sphere3 Hologram of 3D Point Cloud of 3 spheres (polished gold, pearl and silver) with white light from West.
  • Resolution: 1920x1080
  • Pixel pitch: 8um
  • Wavelength: 633, 532 and 460 nm
  • Depth range of object: 0.2783 - 0.2817 m
  • Reconstruction distance: 0.2784 m
  • Point cloud: Download!
Download!
Size: 45 MB
Venus Hologram of a point cloud of a head statue of Venus
  • Resolution: 2048x2048
  • Pixel pitch: 8um
  • Wavelength: 633, 532 and 460 nm
  • Depth range of object: 0.2955 - 0.3045 m
  • Reconstruction distance: 0.2957 m
  • Point cloud: Download!
Download!
Size: 90 MB
Biplane16K Hologram of 3D point cloud of same dice in 3 different depths.
  • Resolution: 16384x16384
    Pixel pitch: 1um
  • Wavelength: 633, 532 and 460 nm
  • Depth range of object: 0.037 - 0.049 m
  • Reconstruction distance: 0.037 m
  • Point cloud: Download!

Point cloud dataset Biplane is courtesy of ScanLAB Projects.
Update: January 2020, correct color dispersion
Download!
Size: 5.99 GB

Release: January 2018

CGH Dataset: Interfere-IV with WUT

Fourier holograms

This dataset consists of 4 computer-generated and 4 optically recorded Fourier holograms, in collaboration with Warsaw University of Technology.
It accompanies our submitted journal paper. If you report results of this benchmark, we request that you cite our work.
The full set of the test holograms along with their acquired quality scores and other related data to these experiment are publicly available at: http://ds.erc-interfere.eu/holodb/

Fourier holograms extra: DeepChess

Additional hologram like CGH "Chess" but with twice its depth. For more documentation on the generation and on how to reconstruct see INTERFERE-IV link above.
  • Resolution: 2048x16384
    Pixel pitch: 3.45um
  • Wavelength: 532 nm
  • Depth range of object: 0.3909 - 1.6063 m
  • Reconstruction distance: 0.3964 m
  • Update: June 2020 Download!
    Size: 244 MB

    Release: September 2019

CGH Dataset: Interfere-V

Fresnel holograms

This dataset consists of computer-generated using a novel Ray-traycing technique.
The dataset will be continually updated in the future and the CGH method will be published.

CornellBox2 8K Hologram from the CornellBox. Requires propagation further away from the scene, before perspective reconstruction is possible.
  • Resolution: 8192x8192
    Pixel pitch: 3um
  • Wavelength: 633 nm
  • Depth range of object: 0 - 0.0245 m
  • Reconstruction distances of interest: 0.004, 0.012, 0.015, 0.0245 m
July 2020
Download!
Size: 479 MB
CornellBox2 10K Hologram from the CornellBox. Generated from the 8K version by angular spectrum propagation. Suitable for <=4096px aperture perspective reconstruction.
  • Resolution: 10240x10240
    Pixel pitch: 3um
  • Wavelength: 633 nm
  • Depth range of object: 0.210 - 0.235 m
  • Reconstruction distances of interest: 0.214, 0.222, 0.225, 0.2345 m
Update: July 2020
Download!
Size: 742 MB
CornellBox3 16K Hologram from the CornellBox. Suitable for <=4096px aperture perspective reconstruction.
  • Resolution: 16384x16384
    Pixel pitch: 2um
  • Wavelength: 633 nm
  • Depth range of object: 0.210 - 0.286 m
  • Reconstruction distances of interest: 0.220, 0.228, 0.25, 0.269, 0.28615 m
Update: August 2020
Download!
Size: 1944 MB
DeepCornellBox 16K Hologram from a variant of CornellBox. (10x as deep as wide; populated with more objects). Suitable for <=4096px aperture perspective reconstruction.
  • Resolution: 16384x16384
  • Pixel pitch: 2µm
  • Wavelength: 633nm
  • Depth range of box: 0.209-0.5304m
  • Reconstruction distances of interest: 0.2090 0.2122 0.2270 0.2500 0.2550 0.3013 0.3425 0.3639 0.4051 0.4080 0.4183 0.4446 0.4891 0.5304 m
  • For detailed list of all objects focused in their center, see zLlong in mat file.
Update: April 2021
Download!
Size: 3,84 GB
CornellBox4 16K Hologram from a variant of CornellBox. (10x as deep as wide; populated with more objects). Suitable for <=4096px aperture perspective reconstruction.
  • CornellBox4 16K
  • Like CornellBox3 but with more natural contrast normalization.
  • Pixel pitch: 2µm
  • Wavelength: 633nm
  • Depth range of object: 0.210-0.235m
  • Reconstruction distances of interest: 0.224, 0.228, 0.25, 0.28615 m
Update: April 2021
Download!
Size: 3,85 GB

Release: September 2019

Media file: Visualization of computer-generated holograms on lightfield displays

This media file accompanies the following publication:
A. Symeonidou, D. Blinder, B. Ceulemans, A. Munteanu, and P. Schelkens, "Three-dimensional rendering of computer-generated holograms acquired from point-clouds on light field displays," Proc. SPIE 9971, Applications of Digital Image Processing XXXIX, 2016

To view the video showing the three-dimensional rendering of computer-generated holograms acquired from point-clouds on light field displays, press here. View video!
For more information, you can contact Athanasia Symeonidou (asymeoni(AT)etrovub.be).

Release: August 2016

MATLAB code: Efficient Multiscale Phase Unwrapping Methodology with Modulo Wavelet Transform

This MATLAB code implements a methodology that can accelerate existing phase unwrapping algorithms using a multiscale approach. Two algorithms have currently been accelerated:
- M. A. Herraez, D. R. Burton, M. J. Lalor, and M. A. Gdeisat, "Fast two-dimensional phase-unwrapping algorithm based on sorting by reliability following a noncontinuous path," Appl. Opt. 41, 7437–7444, 2002
- J.M. Bioucas-Dias, G. Valadao, "Phase Unwrapping via Graph Cuts," in Image Processing, IEEE Transactions on, vol.16, no.3, pp.698-709, 2007

The MATLAB code is freely available for download here. Download!
For more information, you can contact David Blinder (dblinder(AT)etrovub.be).

Release: June 2016

MATLAB code: Studies on the sparsifying operator in compressive digital holography

This MATLAB code generates the figures in the manuscript entitled "Studies on the sparsifying operator in compressive digital holography".
The MATLAB code is freely available for download here. Download!
For more information, you can contact Stijn Bettens (sbettens(AT)etrovub.be).

Release: March 2017

MATLAB code: Bounds and Conditions for Compressive Digital Holography

This MATLAB code generates the figures in the manuscript entitled "Bounds and Conditions for Compressive Digital Holography".
The MATLAB code is freely available for download here. Download!
For more information, you can contact Stijn Bettens (sbettens(AT)etrovub.be).

Release: March 2017

MATLAB code: Sparseness Significance Ranking Measure (SSRM)

This media file accompanies the following publication:
A. Ahar, A. Barri, P. Schelkens, "From Sparse Coding Significance to Perceptual Quality: A New Approach for Image Quality Assessment," IEEE Transactions on Image Processing, 2017
The MATLAB code is freely available for download here. Download!
For more information, you can contact Ayyoub Ahar (aahar(AT)etrovub.be).

Release: November 2017

Please do not forget to cite our work if you use any of the content provided.

Need more information?

Contact us and we'll get in touch the soonest possible.