A True Volumetric Three Dimensional Display


US Patent 5,801,666



The visualization of three (or higher) dimensional information is a general problem which, through the years, has achieved considerable attention. On the most basic level, the world around us is (at least) three dimensional; a two dimensional representation of it is a compromise. A true three dimensional display will provide an unprecedented level of realism which will fold back many of the barriers which prevent our understanding of spatial complexity.

Recently we have been developing a true volumetric monitor for the direct display of three dimensional information. In our work so far, we have designed, built and tested several prototypes that successfully prove the operating principles of our concept. To date, we have built a fiber based working prototype with 76,000 voxels, and an integrated waveguide based prototype with 2000 emitters. Our long term aim is, in conjunction with private industry, to develop a commercially competitive device that may be interfaced as an additional graphics display monitor to a high end personal computer or network workstation. In all cases our choice of materials and our proposed architecture bode well for the economical manufacturability of this innovation.

The general concept for this three dimensional monitor is a fairly straightforward extension of a two dimensional screen composed of an array of picture-elements, or "pixels." Consider here, a three dimensional stack of pixels (actually, volume-elements, or "voxels" becomes the appropriate phrase for such a three dimensional fabric) which, when off, are completely transparent. When addressed, a voxel becomes optically active and emits light. In this way, a three dimensional pattern may be directly built up from a set of activated voxels.

The particular implementation of the voxel is a passive scattering or fluorescing agent. Each voxel is pigtailed to a dielectric waveguide that routes light to the voxel. We have built a large scale display measuring 0.6 cubic meters with 76,000 voxels using optical fibers as the waveguide. We have also built a very high resolution 140 cubic centimeter display with ion exchanged integrated waveguides. To eliminate stray Fresnel reflections off the different glass-to-air surfaces that may distort an image, we fill the voids between the voxels with an index matching medium. Scaling calculations based on accumulated Fresnel scattering indicate that, with reasonable care in design, quality assurance and cleanliness, monitors with more than 10 million voxels layers are possible.

The waveguides terminate in a carefully designed array, which matches a spatial light modulator that acts as a switching array. An attractive solution for this switching array, which we have used so far, is a liquid crystal display, similar to the type currently used in laptop computers. A single light source, currently an array of halogen bulbs provides the input light power. With this illumination/addressing system, our display is highly parallel, and highly standardized to current technology. We run at video rates off a pentium computer. Further, our three dimensional monitor may be made interactive through this switching network.


In constructing an application list for this three dimensional display it is important to recall the relationship between the observer and the display. The goal of most domes or head-mounted displays is the immersion of the operator into a three dimensional scene. In contrast, the device being developed here allows the viewer, especially groups of viewers, to observe --- in a quantitatively true way --- an artificial (or real) three dimensional scene from many different, external, vantage points. Given this, a tremendous array of applications may be thought of for this volumetric display; any image may be projected in true three dimensionality. In this discussion, we divide our application areas into four rough markets: Scientific and Engineering; Commercial; Medical; and Military.

As engineers and scientists, it is natural for us to identify a multiplicity of technical design and visualization tasks that will be eased with this volumetric display. The layout of any mechanical or architectural form and the relation of one part to another will be imediately and inexpensively realized. Complex assembly processes will e tested virtually, and not by trial and error. Complex three dimensional chemicals and biological structures will be visualized and sterically engineered with the help of this display. Complex mathematical surfaces will be easily displayed, for quantitative systems analysis, or aesthetic appreciation. Scientific visualization and flow field analysis are general problems which also demand such a monitor. Implicit in these examples is the projection of four (or higher) dimensional data onto a three dimensional fabric.

Some of the biggest commercial markets may well be in entertainment. Three dimensional games and films should be very much in demand by our modern society.

There are a variety of medical uses for this three dimensional monitor. The output of a variety of imaging techniques, Magnetic Resonance Imaging, for example, is a three to four dimensional array of data. The monitor described here is ideal for the visualization of tomographic data from these techniques, and hence will greatly ease diagnosis and treatment planning. Since the monitor may be made interactive, a surgical procedure may be practiced and optimized on a computer before an actual operation, or be used in the training of medical students and interns. Alternatively, the display could significantly enhance the effectiveness of remote surgery through telepresence. As an outgrowth of the scientific visualization discussed above, this display will aid in the understanding of molecules, viruses and genes and thus in the engineering of novel drugs which may work through steric factors.

A particularly important military application area for our device lies in the briefing and debriefing of multiple viewers in a combat or training scenario. For example, the training of pilots in combat and tactics requires --- and lacks --- a large, affordable, wide field of view display system that supports air-to-air and air-to-ground engagements. A pilot's situational awareness involves his knowing the relative positions among the lead and the wingman, and the threats and targets. Even though this data may come form radar, FLIRs, or other electronic sensors, a visual display of this information is the most efficient for the pilot to assimilate it. It is difficult to adequately represent multiple aircrafts at multiple altitudes on a fundamentally two dimensional display.