Rendering Interactive Holographic Images

Mark Lucente and Tinsley A. Galyean
MIT Media Laboratory

Published in: Proc. of SIGGRAPH 95 (Los Angeles, CA, Aug. 6-11, 1995). In Computer Graphics Proceedings, Annual Conference Series, 1995, ACM SIGGRAPH, pp. 387-394.

Author's notes:

This is an incomplete HTML version of this paper, provided as an introduction. Complete versions are available in Postscript format and a (Unix) compressed Postscript format.
To supplement the References at the end of this document, additional references regarding holovideo are available.

ABSTRACT

We present a method for computing holographic patterns for the generation of three-dimensional (3-D) holographic images at interactive speeds. We used this method to render holograms on a conventional computer graphics workstation. The framebuffer system supplied signals directly to a real-time holographic ("holovideo") display. We developed an efficient algorithm for computing an image-plane stereogram, a type of hologram that allowed for several computational simplifications. The rendering algorithm generated the holographic pattern by compositing a sequence of view images that were rendered using a recentering shear-camera geometry. Computational efficiencies of our rendering method allowed the workstation to calculate a 6-megabyte holographic pattern in under 2 seconds, over 100 times faster than traditional computing methods. Data-transfer time was negligible. Holovideo displays are ideal for numerous 3-D visualization applications, and promise to provide 3-D images with extreme realism. Although the focus of this work was on fast computation for holovideo, the computed holograms can be displayed using other holographic media. We present our method for generating holographic patterns, preceded by a background section containing an introduction to optical and computational holography and holographic displays.

KEYWORDS

Keywords: Electro-Holography, Holovideo, Computer-Generated Holography, Accumulation Buffer.

CR Categories: I.3.1 [Computer Graphics]: Hardware Architecture - three-dimensional displays, graphics processors; I.3.7 [Computer Graphics]: Three-Dimensional Graphics and Realism.

1. INTRODUCTION

The practical use of three-dimensional (3-D) displays has long been a goal in computer graphics. Three-dimensional displays are generally electronic devices that provide binocular depth cues, particularly binocular disparity and convergence. (See the Glossary on the next page.) Some 3-D displays provide additional depth cues such as motion parallax and ocular accommodation. The reference by McKenna[1] contains a good discussion of the visual depth cues and a detailed evaluation of 3-D display techniques. Recently, some 3-D displays have been used interactively. A 3-D display allows the viewer to more efficiently and accurately sense both the 3-D shapes of objects and their relative spatial locations, particularly when monocular depth cues are not prevalent in a scene. When viewing complex or unfamiliar object scenes, the viewer can more quickly and accurately identify the scene contents. Therefore, 3-D displays are important in any application involving the visualization of 3-D data, including telepresence, education, medical imaging, computer-aided design, scientific visualization, and entertainment. The merit of a 3-D display depends primarily on its ability to provide depth cues and high resolutions. The inclusion of depth cues - particularly binocular disparity, motion parallax, and occlusion - increases the realism of an image. Holography[2] is the only imaging technique that can provide all the depth cues[1]. All other 3-D display devices lack one or more of the visual depth cues. For example, stereoscopic displays do not provide ocular accommodation, and volume displays cannot provide occlusion. Image resolution and parallax resolution are also important considerations when displaying 3-D images. While most 3-D displays fail to provide acceptable image and parallax resolutions, holography can produce images with virtually unlimited resolutions. In optical holography, a recorded interference pattern reconstructs an image with an extremely high degree of accuracy. A holographic pattern - called fringes - can be computed and used to generate a 3-D image, most recently in real time[3]. Both the computing process and the displaying process are significantly more difficult than in other 3-D display systems. Nevertheless, a real-time electro-holographic ("holovideo") display can produce dynamic 3-D images with all of the depth cues and image realism found in optical holography. Therefore, holovideo has the potential to produce the highest quality 3-D images. Also, to view holographic images, the viewer is unencumbered by equipment such as glasses or sensors. Figure 1 illustrates the basic functionality of holovideo. Figure 1: A typical real-time 3-D holographic (holovideo) display. When positioned in the viewing zone, viewers see a 3-D image in the vicinity of the output aperture. The input signals are generated by a computer. The size and complexity of holographic fringe patterns often precludes their computation at interactive rates. In the field of computational holography[4], a discretized holographic fringe pattern is generated by numerically simulating the propagation and interference of light. Typical sampling sizes are smaller than the wavelength of visible light. Therefore, a computer-generated hologram (CGH) must contain a huge number of samples. Furthermore, the cost of calculating each sample is high if a conventional approach is taken. Even with the power currently available in scientific workstations, researchers in the field of computational holography commonly report computation times in minutes or hours. In recent work[5], Lucente used a massively parallel supercomputer to calculate holographic fringes at interactive rates. In this paper we describe the first use of a standard graphics workstation to render and display holograms at interactive rates. The graphics workstation provided both a platform for generating image information and the computing power for generating holographic fringes. Our use of a computer graphics workstation also eliminated data transfer bottlenecks that often prohibit interactive computation. We believe that new holographic displays will continue to emerge in the future, necessitating rapid rendering (computation) of holograms. Our work is an important first step toward practical computation systems for holovideo. The following Background section gives a brief summary of the principles of holography. The computer generation of holograms and their display are reviewed, including a brief description of the real-time holographic display that we used in this research. In the section Method for Rendering Hologram we describe the hologram rendering algorithm, including the initial processing of object scene data to provide for realistic lighting, shading, occlusion, and other pictorial depth cues. Finally, we present results, future work, and a conclusion.

GLOSSARY

binocular disparity: the binocular depth cue effected by the slight differences between the two retinal images seen by each eye. The depth sensation caused by binocular disparity is called stereopsis.
convergence: a binocular depth cue effected when the eyes rotate to align the retinal images seen by each eye.
occlusion, overlap: monocular depth cue effected when one part of image is obstructed by another overlapping part.
ocular accommodation: a monocular depth cue in which the eye senses depth by focusing at different distances.
parallax, motion parallax: a (monocular) depth cue sensed from the apparent change in the lateral displacements among objects in a scene as the viewer moves. A display which provides parallax allows the viewer to move around the object scene.
pictorial depth cues: the monocular depth cues found in 2-D images, including occlusion, linear perspective, texture gradient, aerial perspective, shading, and relative sizes.
3-D Display Types
stereoscopic: a 3-D display type that presents a left view of the imaged scene to the left eye and a right view to the right eye. Examples include boom-mounted, head-mounted, and displays using polarizing glasses.
autostereoscopic: a 3-D display type that presents left and right views of the imaged scene without special viewing aids. Examples include lenticular, parallax barrier, slice stacking, and holography. Some provide motion parallax by presenting more than two views.
holovideo: a real-time 3-D electro-holographic display.
Additional Terms
basis fringe: an elemental fringe pattern computed to diffract light in a specific manner. The name "basis fringe" is an analogy to mathematical basis functions. Linear summations of basis fringes were used as holographic patterns.
computational holography, computer-generated holography: the numerical synthesis of holograms.
hololine: a horizontal line of samples of a holographic fringe pattern.
horizontal-parallax-only or HPO: possessing horizontal parallax but not vertical parallax. The viewer sees the same vertical perspective of the imaged scene regardless of the vertical location of the eyes.
image volume: the volume occupied by a 3-D image.
image resolution: the number of resolvable image features in the lateral dimensions of an image.
parallax resolution: the number of different perspective views available to the viewer.

REFERENCES

[1] M. McKenna and D. Zeltzer. Three dimensional visual display systems for virtual environments. Presence: Teleoperators and Virtual Environments. Vol. 1 #4, 1992, pp. 421-458.

[2] P. Hariharan. Optical Holography: Principle, Techniques and Applications. Cambridge University Press, Cambridge, 1984.

[3] P. St. Hilaire, S. A. Benton and M. Lucente. Synthetic aperture holography: a novel approach to three dimensional displays. Journal of the Optical Society of America A, Vol. 9, #11, Nov. 1992, pp. 1969 - 1977.

[4] W. J. Dallas. Topics in applied physics. In B. R. Frieden, editor, The Computer in Optical Research, volume Vol. 41, chapter 6: "Computer-Generated Holograms", Springer-Verlag, New York, 1980, pages 291-366.

[5] Mark Lucente. Interactive computation of holograms using a look-up table . Journal of Electronic Imaging, Vol. 2, #1, Jan 1993, pp. 28-34.

[6] F. Mok, J. Diep, H.-K. Liu and D. Psaltis. Real-time computer-generated hologram by means of liquid-crystal television spatial light modulator. Optics Letters, Vol. 11 #11, Nov. 1986, pp. 748-750

[7] S. Fukushima, T. Kurokawa and M. Ohno. Real-time hologram construction and reconstruction using a high-resolution spatial light modulator. Applied Physics Letters, Vol. 58 #8, Aug. 1991, pp. 787-789.

[8] S. A. Benton. ``Survey of holographic stereograms''. In Processing and Display of Three-Dimensional Data, Proceedings of the SPIE, volume 367, pages 15-19, Bellingham, WA, 1983.

[9] Michael Halle. The Generalized Holographic Stereogram. Master's thesis, Massachusetts Institute of Technology, 1991.

[10] Joseph W. Goodman. Introduction to Fourier Optics. New York: McGraw-Hill. 1968.

[11] J. R. Fienup. Iterative Method Applied to Image Reconstruction And to Computer-Generated Holograms. Optical Engineering, Vol. 19 #3, May/June 1980, pp.297-305.

[12] Mark Lucente. Diffraction-Specific Fringe Computation for Electro-Holography. Ph.D. Thesis, Massachusetts Institute of Technology, 1994.

Acknowledgments:

Silicon Graphics Inc. made the wonderful graphics hardware that was used in this paper.

And we thank our many sponsors.