摘要:
A device for reproducing a stereoscopic image viewable within an increased observation area, the device including an image display device which displays a plurality of images of an object pixel by pixel from various perspectives, a polarization arrangement for producing a first polarized image associated with a right eye and a second polarized image associated with a left eye of an observer, a lens network spaced directly after the polarization arrangement which directs beams associated with each of the left and right polarized images to the respective left and right eye for representing the object pixel by pixel for observing the stereoscopic image, wherein the lens network and polarization arrangement are arranged such that the stereoscopic image can be perceived in a first region of the observation area without polarized spectacles, and in a second region outside the first region of the observation area, by means of polarized spectacles over the left and right eyes.
摘要:
A system and method that determines a number of scan lines for each image of a view using a resolution of the recording media, a pitch of the lenticules and a number of views needed to minimize angular transitions between views. The viewing range is also increased by allowing the primary visual field to be offset with respect to the projection field of the lenticules such that the image lines can be positioned under lenticules adjacent to the lenticule projecting the image lines as the distance from a central viewing position increases.
摘要:
A stereoscopic endoscope is provided having a primary optical system for forming an image of an object, a device for dividing the light transmitted by the primary optical system into two light beams, and a pair of optical systems for detecting the image formed by the primary optical system. Each of the secondary optical systems may receive one of the light beams and form an image of the object on a corresponding imaging device. The stereoscopic endoscope may also include a device for adjusting a position of the light dividing device relative to an optical axis and an exit pupil of the primary optical system, and a device for detecting a position of the light dividing device in accordance with each of the output image signals.
摘要:
Several different methods alone or in combination can be used to reduce image jump or stutter. The image plane on which the original image is taken can be mechanically or electronically rotated about a point of rotation in the scene. The number of original images which are viewed could be increased by interpolation. The subtended viewing angle of the viewpoints can be reduced and viewpoints within the subtended viewing angle can be increased by interpolation. Reduced subtended viewing angle images can be created for each eye. The depth of focus of the images can be changed by identifying the images that jump and blurring the images outside the area of interest. The original images can be processed to move the image plane further back from the objects in the scene. Some of the images ordinarily existing between the right and left eye perspectives can be eliminated. The objects in the background of the scene being photographed can be identified and shifted forward in the scene. The highest detail volume of the images can be identified and the images can be processed to rotate around the volume. The highest detail area can be identified and the original images processed to displace the images to minimize the movement of the highest detail area. The original images can be viewed and a point within each image identified to align the point from view to view using a linear or other transformation.
摘要:
The invention relates to a single-camera autostereoscopic picture-taking device implementing an array of cylindrical lenses, characterized in that it comprises in succession: a single inlet objective (L1, L2); said lens array (20), which has a focal length such that for an image area equal to the pitch (p) of the lenses making it up, the image of the entrance pupil of the telecentric objective (L1, L2) has a nominal width equal to said pitch; a transfer optical system (L3, L4) having magnification of less than unity; and an image sensor (22), the transfer optical system (L3, L4) directing the light rays that emerge from the lens array (20) onto the image sensor (22), the image (21) of the lens array (20) in the transfer optical system (L3, L4) being such that the pitch (p) of the lenses of the lens array (20) corresponds therein to an integer number of image points (pixels) of the image sensor (22).
摘要:
A display concept designed to interface with source material from a computer, video recorder or video camera. The interface to a computer generates a series of images that are separate viewpoints of a desired scene to be viewed in three dimensions. The images are based upon a horizontal translation of viewpoints. In the case of a video camera, the interface captures video images from multiple viewpoints and delivers these images in a successive manner to a location at which they can be viewed. The camera requires a single lens and sensor instead of multiple lenses. In the video recorder case, an interface stores images from a computer or video camera and plays back images in the proper sequence for the display. A display system can also operate as a film recorder to provide hard copy, three-dimensional images. A computer, video recorder or video camera thus provides images to a surface which can be viewed and which can present two-dimensional images which are changeable into three-dimensional images. In this way, the output from a computer, video recorder or a video camera can be used to store images and to play back the images in the proper sequence for a three-dimensional display.
摘要:
The improved method and apparatus for recording a stereoscopic image read a plurality of original images photoelectrically, split each original image into linear images of a desired pitch, perform any necessary image processing operations such as arranging the linear images in the recording order, and record the linear images by scan exposure on a lenticular light-sensitive material having a lenticular sheet and a recording layer. The method and apparatus permit the lenticular light-sensitive material to be scan exposed from the rear side and they also enable the linear images to be recorded with increased line width and a stereoscopic image to be formed with satisfactory registration between the principal objects of the original images. The improved lenticular light-sensitive material has an anti-halation layer provided between the lenticular sheet and the recording layer, thus permitting scan exposure to be effected from the rear side. According to the present invention, stereoscopic images of high quality that have no perceivable unnaturalness such as flickers can be recorded with great ease and at any desired magnification.
摘要:
A stereoscopic image of an object is focused on an image pickup device based on the degree to which left and right images of the object are coincident on the image pickup device. Signals produced by the image pickup device that represent the left and right images are compared with each other to develop a control signal that indicates the coincidence of the left and right images on the image pickup device, and the spacing between the image pickup device and an optical element that directs the images onto the device is changed in response to the control signal to increase the coincidence. Two alternative types of comparison are performed, each of which indicates the degree of coincidence (and hence the quality of focus) of the left and right images on the image detector. In one approach, the left and right images are compared to determine their cross correlation on the image pickup device, and the spacing is changed to maximize the correlation. Alternatively, the comparison determines the difference between the left and right images on the image pickup device, and the spacing is changed to minimize the difference. The focusing technique is useful in, e.g., the optical system of an endoscope. Rangefinding and videometry are also performed to determine the distance to the object being viewed, as well as its position and size.
摘要:
The improved method and apparatus for recording a stereoscopic image read a plurality of original images photoelectrically, split each original image into linear images of a desired pitch, perform any necessary image processing operations such as arranging the linear images in the recording order, and record the linear images by scan exposure on a lenticular light-sensitive material having a lenticular sheet and a recording layer. The method and apparatus permit the lenticular light-sensitive material to be scan exposed from the rear side and they also enable the linear images to be recorded with increased line width and a stereoscopic image to be formed with satisfactory registration between the principal objects of the original images. The improved lenticular light-sensitive material has an anti-halation layer provided between the lenticular sheet and the recording layer, thus permitting scan exposure to be effected from the rear side. According to the present invention, stereoscopic images of high quality that have no perceivable unnaturalness such as flickers can be recorded with great ease and at any desired magnification.
摘要:
A 3D camera for acquiring and recording a plurality of 2D images at different viewing angles without relocating the camera wherein a large-aperture taking lens is used for image forming and an aperture controlling means is used to select the section of the lens' aperture through which light rays are transmitted to form a 2D image at one viewing angle at a time. The 3D camera can also have a mirror assembly to direct light rays traversing different sections of the lens' aperture to form 2D images on different image planes.