SAN JOSE, CA--Photonics West was the site for the unveiling of a new three-dimensional (3-D) display technology by Floating Images Inc. (Westbury, NY). The technology allows viewers to see full-parallax 3-D images that appear to float in free space, but without the need for special glasses. Not only are the images quite compelling, but they can be created in a variety of implementations and for very low cost. Computer gaming, location-based entertainment, and business and financial centers are all targets for this technology.
One of the most effective demonstrations of the device used Floating Image`s large free-space imaging system. With it, a bust of Beethoven was suspended in thin air, rotated, and moved back and forth. In many ways, it is very similar to the 3-D effect experienced during Disney`s Honey I Shrunk The Audience entertainment adventure, except no glasses are required.
The 3-D image allows viewers to move their heads and obtain a different view of the object (horizontal and vertical binocular parallax) while providing traditional depth cues such as perspective and image occlusion. According to Floating Images president Gene Dolgoff, the realism of the 3-D generation lies in the use of depth disparity. "If your eyes focus at different distances, the brain interprets that as a depth cue," explained Dolgoff. "Consequently, we produce images that actually are at different depths. This technique produces no eye strain or headaches, which can happen with stereoscopic or autostereoscopic systems."
Concepts simple
Surprisingly, the patented concepts behind Floating Images devices are quite simple. Ordinary television or computer monitors can be used and simply fitted with an optical adapter. "We plan to sell these devices, which we call Real-Depth Adapters, for less than $70," added Dolgoff.
What Dolgoff and coinventor Louis Tullo discovered was that separating a flat two-dimensional (2-D) image into only two image planes and adding some traditional depth cues was sufficient to produce very good 3-D images. A standard TV or PC image can be reformatted into foreground and background scenes and displayed on the top and bottom half of the display (see figure on p. 24). A 45° mirror and beam combiner are used to produce a virtual image of the background scene directly in back of the foreground scene. The result is two discrete and easily discernable image levels. Objects or figures can navigate between these two levels giving the illusion of movement in 3-D space.
Applications
Having even two discrete image levels has several benefits and potential applications. For example, many entertainment games feature a character in a foreground scene moving against a background scene. Explains Andrew Johnston, president of WizBang! Software Productions (Mercer Island, WA), the first video-game development company licensed to use the product, "It is very simple to split the memory buffer and separately render the background and the foreground images. Games could ship with both the standard and Real-Depth-enabled versions. Only a single keystroke is needed to activate the Real-Depth mode."
Another application for this discrete level approach is in the business arena. The "Floating Windows" concept allows users to create two Windows sessions in the top and bottom half of the display. Using the Real-Depth Adapter, the two applications, such as a spreadsheet-analysis application and a stock-market tracking program, are superimposed on to¥of each other, but at different depth locations to eliminate confusion. The spreadsheet foreground can be made transparent so the user can move between applications by simply refocusing on the background image.
Kiosks or business presentation models would employ two Fresnel lenses to float imagery into free space. Such devices produce 3-D images that appear to lea¥out of the imager.
Many applications, however, will require continuous motion rendering between these two image planes. The company is now working on concepts to achieve this capability. On display at Photonics West were several animated clips featuring various approaches.
Clearly, some methods were better than others, and the quality often depended upon the type of imagery being used. For example, one approach is to use electronic transforms to separate image areas of high spatial frequency from areas of lower spatial frequency. Low-spatial-frequency areas get mapped to the background plane while higher-frequency components ma¥to the foreground scene. Additional algorithms are needed to manage the movement of objects within this space. The goal is to convert conventional 2-D source material to the Real-Depth format in real time.
Floating Images is now developing the next-generation prototype systems and working with software developers in several industries. Real-Depth-enabled games should be available by Christmas 1998.