VISION III IMAGING, INC.

Vision III Imaging, Inc. (the Company) specializes in advanced imaging technology and is based in Vienna, Virginia. The Company has developed the v3¨ parallax scanning technology for recording visual parallax information (i.e. differing points of view of the subject).

v3 parallax scanning creates sharper and more realistic images with enhanced textures, shape definition, and improved quality. The enhancement is accomplished by exploiting certain psychophysical mechanisms of the human visual system and in particular, those visual mechanisms used in depth mapping.

The v3 technology can be used to capture or create imaging for conventional television programming, high definition video, 35 mm feature films, video games, computer animation, Internet pictures and mobile imaging. v3 images can be recorded and projected on motion picture film, broadcast to home televisions, streamed as Internet video, displayed on mobile devices, or created by computer animation or games. No glasses or special display equipment are required. All the viewer need do is look at the image. v3 technology can also be

applied to the vast of array of medical, industrial, scientific, military imaging.

v3 is not an "either or" acquisition technology. It can be turned up, down, or off. When the scan is turned off, the images appear as normal "flat" images. You can use a little, a lot, or none at all. v3 is both an image enhancement and unique creative imaging tool.

v3 TECHNOLOGY ATTRIBUTES

v3 products have been developed as both hardware and software tools that allow parallax visualization in a variety of different mediums. All of the tools create enhanced images with the following attributes:

Integrates into current imaging workflow

Renders object shape and terrain features obvious

Provides operator "value added" software tools

Provides viewer a "value added" experience

Allows for seamless image distribution across a network or via broadcast

No increase in bandwidth

Boosts operator recognition performance

Enhances threat, change and contraband detection

Provides data for object mensuration and identification

Near real time depth enhancement processing of payload sensor images

Real-time depth enhancement of DPS equipped lenses

Post image analysis and change detection using Parallax Image display (PID) software tool set

Parallax visualization of geographical browser and value added data

EXPLOITATION OF HUMAN PERCEPTION MECHANISMS

In his 1995 book, Foundations of Vision, Brian Wadell states, "Perception is an interpretation of the retinal image, not a description. Information in the retinal image may be interpreted in many different ways. Because we begin with ambiguous information, we cannot make deductions from the retinal image, only inferences. ....we have learned that the visual system succeeds in interpreting images because of statistical regularities present in the visual environment and hence in the retinal image. These regularities permit the visual system to use fragmentary information present in the retinal image to draw accurate inferences about the physical cause of the image. For example, when we make inferences from the retinal image, the knowledge that we live in a three dimensional world is essential to the correct interpretation of the image. Often, we are made aware of the existence of these powerful interpretations and their assumptions when they are in error, that is, when we discover a visual illusion."

While it is not possible to create a stereo image on a standard display without special equipment, it is possible to create an image with enhanced texture and depth. The v3 parallax scanning technology creates moving images with enhanced texture and depth on standard displays (television, movie screens and computer monitors) without the necessity of special screens or the use of viewing glasses. v3 images can be recorded on normal film, videotape or digitally using industry standard systems. The v3 image enhancement is accomplished entirely by a lens or synthetically generated in software using the patented parallax scanning technology.

PSYCHO-PHYSICAL RESEARCH

It is known that the act of visual perception is a cognitive exercise and not merely a stimulus response. In other words, perception is a learned ability which we develop in infancy. Kenneth Ogle of the Mayo Clinic, reported 1967 that left and right-eye information can be presented alternatively to the left and right eyes, resulting in depth perception as long as the time inter val does not exceed 100msec. Visual researcher David Marr has suggested that perceptual fusion of binocular information occurs in a short-term memory buffer by means of some sort of visual depth mapping. In 1984, Edwin Jones of the University of South Carolina reported that the human brain can accept and process parallax information without regard to the direction of the parallax, i.e. horizontal, diagonal or vertical. A. P. McLaurin of the University of South Carolina, has stated that if visual information is in fact compared in a temporary memory and does not have to be received simultaneously, there is no reason why stereoscopic information that is appropriately sequenced at the proper rate cannot be observed by the single eye.

In August ‘98, the University of Virginia - Cognitive Science Department received an Innovation Award from the Virginia Center for Innovative Technology (CIT) to fund a research project to study the perceptual aspects of parallax scanning on the human visual system. This project and its subsequent report t

were completed in March ‘99.

The UVA report titled Perceived depth is enhanced with parallax scanning, was the first independent study of the v3 parallax scanning technologies. Dr. Dennis Proffitt and Tom Banton’s work confirm that parallax scanning enhances perceived depth in images, especially when the object depth is large. The more depth in the scene, the more parallax scanning enhances its perception by a viewer on a standard television screen without the aid of special glasses.

PARALLAX SCANNING

v3 methods rely on discrete parallax differences between depth planes in a scene. The differences are caused by a parallax scan. When properly balanced (tuned) and displayed, the discrete parallax differences are perceived by the brain as depth.

A continuously moving parallax scan records a pattern of sequential parallax views on a single strip of film or video tape. The lens’s optical axis sweeps in the plane of the nominal X and Y axes around the nominal optical Z axis, pivoting on the optical convergence point (out along the Z axis), so that it passes through positions having parallax in relation to the optical convergence point. The circular scanning of the lens’s optical axis traces out a coaxial cone pattern with the convergence point as its apex.

Early tests revealed that the brain will translate parallax scanned information into depth information at scanning frequencies of between 3 - 6 Hz, and that the ideal frequency is 4.31 Hz.

As previously stated, a parallax scanning allows live-action or computer generated images to exploit certain depth mapping psycho-physical mechanisms inherent in the human visual system. v3 parallax scanning images create a greater sense of dimension, sharpness and overall realism when displayed by standard motion picture projection, broadcast television, mobile device or computer monitor. The v3 technology is not an either or situation, rather, like the volume knob on a radio, the amount of parallax scanning can be turned up or down. v3 is a creative tool that allows artists to manipulate image texture and dimension.

THE DPS®

A v3 Digital Parallax Scanner (the DPS) mechanism has been designed. This design employs two custom linear actuators and a central pivoting armature that holds the iris. The two parallel linear actuators have coordinated motion in such a way as to produce both x and y motions of the iris. For illustrative purposes think of the way, a tank moves. If both tank treads move forward/backward the "gun tip" moves forward/backward (both treads moving normally). If one tread moves opposite to the other (turning, both treads moving differentially) the "gun tip" would move left or right. It is this type of differential motion that allows the iris to be positioned in any area of the optical axis in the lens, and do it in the smallest possible space.

The linear actuators consist of a moving coil and fixed magnetic yoke assembly, very similar to the typical actuator that controls the read/write heads in a computer hard drive. By incorporating miniature, high-resolution optical encoders, the entire scanner mechanism control system is completely digital.

The iris itself is a conventional 12-leaf design, but with its actuation controlled by a miniature DC motor. The entire mechanism is very small and compact, while being flexible enough to be adapted to a variety of lenses.

The control electronics are divided into two parts. The primary control system controls low-level functions, and is part of the mechanism enclosure that mounts directly on the lens. A common Preston style (cable or wireless) handheld control unit or a laptop computer contains the secondary control system and manage higher, level user interface and storage functions.

The Company introduced the v3 AX3 parallax scanning unit for the Fujinon 18x and 22x or Angenieux 19x and 26x HD zoom lenses. The AX3 utilizes DPS technology packaged for broadcast HD applications.

AX3 HD ZOOM LENS UNIT

The AX3 is designed to be inserted into the lens’ optical path at or near the plane of the iris allowing the capture

of parallax views. HD video shot using AX3 equipped lenses can be recorded using industry standard cameras and broadcast to the home without the necessity of any additional equipment.

The AX3 features conveniently located on-board parallax scan operator controls. The unit software offers function parameter preset features that can be set and easily "triggered" while shooting. Parallax scan tracking is

automatic and changes according to focal length while zooming. Software presets and automatic parallax scan functions can be customized by the operator. Additional remote controls are also available for wireless and

tethered operation.

THE MOE® PRIME LENSES

The Company has developed the v3 Moving Optical Element (MOE) camera lens family that employs DPS technology to inject vivid impressions of depth into visual image sequences (recorded on standard film or video tape). Motion picture film shot using MOE lenses is recorded using industry standard cameras/film and can be projected or transferred to videotape. The MOE lens system contains the following features:

PL-mount compatible (Arriflex, Moviecam, and Aaton)

Digital lens controls

Self calibrating features

Automatic parallax scan control software

Remote control of lens functions (iris, focus & parallax scan) via Preston

Cinema Products F/X controller unit

Random adjustable Image Shaker up to 20 Hz

v3 SYNTHETIC IMAGING

Initial v3 products have been lens systems for motion picture film and HD video cameras. However, Company research and development has resulted in software products for numerous computer graphics applications. In particular, current target markets for v3 computer imaging products are 3D computer (video) games and 3D animation and rendering programs. The two major v3 software imaging products are described below.

v3 GAME GODE

v3 Game Code has been developed incorporating v3 parallax scan methods and control algorithms for use with PC video game software rendering engines. The v3 Game allows a standard 3D game rendering engine to create parallax scanned images on a real-time basis and automatically adjusts scan parameters as the game is played. The v3 features are seamless to the player, but can be manually adjusted if so desired. Games using the v3 Game code function just like normal games, but with the added advantage of being depth-enhanced. There is no significant increase to the game overhead (speed) associated with the v3 Game code and it has been designed to integrate easily into any 3D engine. v3 Game has been written to be compatible with OpenGL and Direct3D, and other proprietary systems. Current game engines tested include Quake (http://www.idsoftware.com/) and Titan (http://talika.fie.us.es/~titan/).

v3 3D COMPUTER ANIMATION

A software plug-in (v3 Plug-in) has been developed to simulate the v3 parallax scanning optical technique using 3D computer animation software packages. Computer animation software package are popularly used for the production of synthetic (computer-generated) images for television advertising, feature films and Internet imaging. The v3 Plug-in allows for parallax scan settings to be applied to the VC in a manner very similar to the automatic control algorithms,that were developed for the v3 Game Code. v3 plug-ins are available for NewTec/LightWave and Silicon Graphics/Maya. Additional plug-ins are in the process of being developed for Avid/SoftImage and Autodesk/3D Studio Max.

CONCLUSION

The potential economic benefits of Vision III’s proposed project in image segmentation are tremendous given the size and variety of the possible markets. Successful project results would advance the state of the art in both

theory and applications. As a commercial technology, image segmentation is still in its infancy. This project’s results could propel the technology into serious commercial products spanning many markets. The applications

reviewed above suggest the enormous array of opportunities, and there would likely be still more that are as yet unforeseen.