Television – Stereoscopic – Picture signal generator
Reexamination Certificate
1997-07-14
2001-12-04
Rao, Andy (Department: 2613)
Television
Stereoscopic
Picture signal generator
C348S043000
Reexamination Certificate
active
06326995
ABSTRACT:
BACKGROUND OF THE INVENTION
When capturing and reproducing 3-dimensional images in the prior art, information from one camera of a stereo pair of cameras was depicted as one color (e.g. orange) or band of colors and information from the other camera of the pair was depicted in a complimentary color or color band. When viewing such images through 3-dimensional viewers, such as red/blue glasses, the reproduced image would not be perceived in color.
The orange elements in the picture are only seen through the blue lens, the red lens “washing out” the orange elements. For the same reason, the green-blue elements are only seen through the red lens. Hence, each eye sees only one of the two colored pictures. But because the different colored elements are horizontally shifted in varying amounts, the viewer's eyes must turn inward to properly view some elements, and turn outward to properly view others. Those elements for which the eyes turn inward, which is what the viewer does to observe a close object, are naturally perceived as close to the viewer. Elements for which the viewer's eyes turn outward are correspondingly perceived as distant. Specifically, if the blue lens covers the viewer's right eye, as is generally conventional, then any blue-green element shifted to the left of its corresponding orange element appears to the viewer as close. The element appears closer the greater the leftward shift. Conversely, as a green-blue element is shifted only slightly leftward, not at all, or even to the right of its corresponding red element, that element will appear increasingly more distant from the viewer.
The above mentioned co-pending applications teach techniques for producing color 3-dimensional images.
When 3-dimensional images are captured, corresponding points of the left image are displaced from the same points in the right image horizontally. A measurement of the amount of displacement is called “disparity”. In the prior art when stereo images are made, the disparity for all subject matter visible in both images is fixed. In digital images, disparity can be measured in terms of the number of pixels an object is displaced in the right image relative to its position in the left image. Fixed focal length lenses are customarily used for the cameras
In an object with zero disparity, the corresponding pixels for the left and right images are perfectly superimposed and the object appears to be located on the screen. Zero disparity objects are seen most clearly when the eyes are crossed just enough to focus on the plane of the screen. Negative disparity objects appear to come out of screen toward the viewer and are seen most clearly when the eyes are more crossed. Positive disparity objects appear to be more distant than the screen and are seen most clearly when the eyes are less crossed.
The eyes cross or uncross in order to get similar image features on or near the fovea of each eye. The “farthest” object that can be seen in an anaglyph is limited by the observers ability to comfortably uncross the eyes. (The usual limit to distant viewing is set by the condition where the eyes look along parallel axes, but such “wall-eyed” condition is rarely comfortable to the observer.)
In an anaglyph, the disparity for all objects is fixed and is measured in terms of pixels of displacement. When one “zooms-in” on a computer image to see more detail, the pixels get larger and the center-to-center spacing between pixels becomes larger. Therefore, constant disparity (measured in pixels) image components become physically farther apart on the screen. In order for the human visual system to fuse image components and produce the sensation of true stereo vision the eyes have to uncross more for each step of “zoom-in”. Eventually, the physical separation between corresponding image components becomes so great that the eyes cannot “uncross” comfortably any more (wall-eyed condition) and stereo depth is lost to the observer.
Some stereo images cover such a great range of depth and will have such widely varying values (even without a “zoom-in”) that some portions of the image will always be out of range of the observer's ability to see the stereo effects, regardless of how the anaglyph was formed.
Three dimensional techniques are closely related to the psychology and physiology of an observer's cognitive processes. Subtle changes in selection of portions of the spectrum presented to each eye can result in significant changes in the observer's perception. Even when viewing the same 3-dimensional image through the same viewers, different observers may perceive a 3-dimensional image in different ways.
The depth location of the point at which the left and right image points for objects at that distance coincided constitutes a “neutral plane” and when observing a fixed disparity 3-dimensional image, the neutral plane would be found at the surface of the medium of reproduction (i.e. paper or CRT display). Items that appear closer than the medium surface and those points in the image which appear behind the neutral plane would have different disparity. The loss of depth perception when disparity exceeds a certain value generally means that when zooming-in on part of a stereo image pair that disparity will become so great that depth perception will be lost. This is a serious drawback when, for example, attempting to use medical images captured in stereo for instructional purposes. Typically, one would need to examine parts of an object in detail by going close up. This problem is analogous to having a fixed focal length microscope and being unable to see close up features which do not lie directly in the focal plane.
Also in the prior art, when capturing 3-dimensional images on film, magnetic tape or the like, there is no way to visually monitor the combined impact of the separate images being captured. As a result there is no way of adjusting disparity or automatically tracking an object and adjusting disparity automatically.
In the prior art, there is no way to control an image so as to position it either in front of or behind a neutral plane in a controllable fashion. This limits the ability to create 3-dimensional animations.
Also in the prior art, there was no way to adjust the views of 3-dimensional images captured on a static medium, such as CD/ROM.
The prior art lacked the ability to zoom-in on portions of a scene when capturing the scene from one location. In order to zoom-in on a scene in the prior art, a stereo camera pair with fixed focal length had to be physically relocated closer to the object being captured.
DISCLOSURE OF THE INVENTION
One advantage of the invention is that it allows for controlling disparity when capturing or reproducing an image.
Another advantage of the invention is that it permits a user to control the disparity by which left and right images are separated.
Another advantage of the invention is the simultaneous adjustment of focal length in stereo camera pairs.
Another advantage of the invention is the ability to adjust camera separation or camera tow-in.
Another advantage of the invention is that it permits zooming-in on portions of a stereo image without losing depth perception.
Another advantage of the invention is the ability to control the location of the neutral plane in 3-dimensional views, thus enabling objects to be controllably placed in front of the neutral plane (popping out of the screen) or behind the neutral plane (in background).
Another advantage of the invention is the ability to create a computer animation using disparity control to produce very realistic animations which move in front of and behind the neutral plane.
These and other objects and advantages of the invention are achieved by providing methods and apparatus for viewing three dimensional images which shift one image view with respect to an other image view to control the amount of disparity between corresponding points of the two views and displays the image views so as to form a three dimensional image. The shifting of one image view with respect to another is accom
McLaine Raymond
Palm Charles S.
McDermott & Will & Emery
Rao Andy
Synthonics Incorporated
LandOfFree
Methods and apparatus for zooming during capture and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Methods and apparatus for zooming during capture and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods and apparatus for zooming during capture and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2600823