JP2007108626A - Stereoscopic image forming system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a stereoscopic image forming system of forming a stereoscopic image having a natural depth. <P>SOLUTION: The stereoscopic image forming system includes: a stereoscopic image pickup means constituted of a lens for imaging an actual image of an object and an image pickup device; and a stereoscopic image display means including an image display device for obtaining a plurality of images in accordance with a distance to an object by changing an image pickup position in accordance with the distance to the object, and then, individually displaying the obtained images different according to distances. By changing a distance up to an image display position in accordance with the distance up to the original position of the object, the stereoscopic image having a depth can be displayed. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a stereoscopic video generation system that captures and displays natural stereoscopic video.

In a conventional stereoscopic image generation system, a right-eye image and a left-eye image are captured by separate cameras, and each image light is displayed with polarized light orthogonal to each other at the time of display. Stereoscopic viewing was realized by wearing polarized glasses that were orthogonal to each other with the left and right eyes, and viewing the right-eye video and the left-eye video separately. (For example, refer to Patent Document 1.) FIG. 18 shows a configuration of a conventional twin-lens stereoscopic camera described in Patent Document 1.
In FIG. 18, 100 is a right-eye camera, 101 is a left-eye camera, 102 is a time-division multiplexed display unit, and 103 is liquid crystal shutter glasses. When shooting a stereoscopic video, the video shot by the right-eye camera 100 and the left-eye camera 101 are alternately rearranged for each field by the time-division multiplexing unit 102 and time-division multiplexed. It is displayed and output alternately. When this is viewed through the liquid crystal shutter glasses 103, the left and right eye liquid crystal shutters are alternately opened and closed in synchronization with the time-division multiplexed left and right eye images, so that the right eye displays only the right eye image and the left eye displays the left eye. By viewing only the eye image, stereoscopic viewing was realized.
Japanese Examined Patent Publication No. 6-34528 (first page, FIGS. 1 and 5)

However, in the conventional configuration, stereoscopic viewing is performed using only binocular parallax, which indicates that the display position of the left and right eye images differs depending on the depth position of the object in the image. The focus position of the eye must be fixed on the multiplexed display unit, and there is a problem that stereoscopic viewing is difficult or fatigued. In addition, there is a problem that glasses must be worn for stereoscopic viewing.
The present invention solves all the above-mentioned conventional problems, and binocular parallax and the focal position change in conformity with the depth position of the object in the image, and further, glasses are not required, so that stereoscopic viewing is possible. An object is to provide a natural stereoscopic image that is easy and does not cause fatigue.

In order to solve the above-described conventional problems, a stereoscopic video generation system according to the present invention includes a stereoscopic video imaging unit including a lens that forms a real image of an object and an imaging device disposed at the real image position. The stereoscopic image capturing means obtains a plurality of images according to the distance to the imaged object by varying the distance from the lens to the imaging position of the imaging device according to the distance to the object, and for each obtained distance. A stereoscopic video display means including a video display device for individually displaying different videos, and the stereoscopic video display means varies the distance to the video display position according to the distance to the original position of the object. Thus, a stereoscopic image with depth is displayed.
With this configuration, the viewer of the image can see a truly stereoscopic image without wearing glasses.

  According to the stereoscopic image generation system of the present invention, the display position of the displayed object image is set according to the distance at which the object actually exists, so that the binocular parallax and the focus adjustment match, and the left and right eyes Can see 3D images in a natural way without the need for glasses because they see the same images.

Embodiments of the present invention will be described below with reference to the drawings.
[Embodiment 1]
FIG. 1 is a configuration diagram of a stereoscopic video generation system according to Embodiment 1 of the present invention. In FIG. 1, reference numeral 1 denotes a convex lens, and 2 denotes an imaging device including a charge coupled device (CCD). In the configuration shown in the figure, if the focal length of the lens 1 is f, the real image of the object D1 located at the distance D in front of the lens is connected to the position behind the lens 1 at a distance D × f / (D−f). Imaged. That is, since the real image position varies depending on the distance D to the object to be photographed, individual images corresponding to the distance of the photographed object can be obtained by photographing while varying the distance from the lens 1 to the imaging position of the imaging device 2. Can be obtained. An image at infinity can be located behind the lens 1 at a distance f, and an image of an object at a distance closer than that can be farther than f. If the distance is set to be D and within a distance of D × f / (D−f), an image of an object at a distance from D to infinity can be taken.

  FIG. 2 shows a detailed structure of the image pickup device 2, and a plurality of transparent image pickup panels 21 are overlapped, and the distance from the lens to the image pickup position is varied by the thickness of each image pickup panel. As the imaging panel, a charge coupled device (CCD) used in a normal camera, or an electrode such as a CMOS sensor using a transparent electrode can be used.

  FIG. 3 is a configuration diagram of the imaging panel 21 and has a transparent electrode 211. The transparent substrate is made of glass or the like, and the semiconductor and the insulating film are transparent because they are formed by injecting a small amount of impurities into silicon which is one of glass materials. The transparent electrode 211 can be made of polysilicon or the like. With this configuration, each imaging panel individually captures images of objects at different distances according to the distance to the lens. In addition, since the plurality of image pickup panels are farther away from the lens, the light is attenuated by the front panel and the amount of light that arrives decreases, so the rear panel increases the sensitivity, or the panel output signal. By amplifying, an image with a constant brightness is obtained regardless of the panel position.

Since the number of image pickup panels must be finite, an image formed at a position where there is no image pickup surface is blurred on the front and rear panels. The amount of blur δ is expressed by assuming that the lens diameter is D, the focal length is f, the distance to the object is d, and the distance from the lens to the imaging panel is f + x.
δ = D | x (d−f) −f ² | / (fd) (1). From this equation, it can be seen that a lens having a small aperture D is good for reducing the amount of blur. Now, when a lens for D = 35.7 mm and f = 50 mm is used and a CCD for HDTV with a width of 35 mm is used as an imaging panel, the resolution is 1920 pixels × 1088 pixels vertically, so the horizontal size of one pixel is ,
35 mm / 1920 = 18.2 μ (2). FIG. 4 shows the amount of blurring of the image on each panel in the case where the imaging panels are sequentially overlapped at an interval of 0.2 mm from the position of the focal length f in terms of the number of pixels. This graph shows that the amount of blur of an object image located in the middle of the panel can be reduced to a certain value or less by overlapping the imaging panels at equal intervals. When the panels are spaced by 0.2 mm, the blur amount of the intermediate video is 4 pixels or less.

  In addition, as shown in FIG. 4, when there are only five imaging panels, when these are arranged at a distance of 0.2 mm from the position of the focal length f of the lens, the distance that can be photographed within 4 pixels of the blur amount It can be seen that the range can be about 2.8 m to ∞. If you want to shoot a closer distance than that, for example, if you place 5 imaging panels far away from the position of the focal length f + 1mm of the lens, you can shoot a distance of about 1.3m to 2.8m under the same conditions. You can see that it can be set. Furthermore, when it is desired to photograph a distance of 1.3 m to ∞ with five imaging panels, the maximum blur amount is doubled, but it can be understood that the imaging panel interval may be increased to 0.4 mm interval. If it is desired to increase the shootable distance range without increasing the amount of blur, the number of imaging panels may be increased.

  In order to make stereoscopic viewing effective, it is desirable that an image of an object at a certain distance is formed on only one panel and not on another panel. For this purpose, it is only necessary to greatly blur out of focus images using a bright lens. Furthermore, in order to remove the remaining unnecessary video, the blurred video does not contain a high frequency component, so the frequency component of the video signal output from the imaging panel is examined, and an area not including a predetermined frequency or more is extracted. If a part is brought close to the maximum luminance value, the part becomes transparent and unnecessary images can be erased. In order to examine the frequency component of the video signal, space-frequency coordinate transformation such as discrete cosine transformation or Fourier transformation may be performed, or the output amplitude of a two-dimensional high-pass filter as shown in FIG. 5 may be examined. good. The filtering process is performed by covering the image data arranged on the plane with the filter shown in the figure and multiplying each coefficient by the luminance value of the pixel at the corresponding position.

In order to make the filtering process more effective, it is desirable to increase the amount of blurring of an unfocused image so as to reliably detect a blurred portion of the image. For this purpose, it is understood from the formula (1) that the lens aperture D should be increased.
In this case, since the depth of focus, which is a distance range in which one image pickup panel is focused, becomes shallow, more image pickup panels are required when shooting a wide distance range with a small amount of blur.
In order to improve the blur of the remaining video after the unnecessary video is made transparent, the high frequency component may be emphasized by passing the video signal through a high pass filter as shown in FIG. If the degree of emphasis is too large, ringing will occur near the edge of the pattern, so adjust it to an appropriate amount.

If the unnecessary image is made transparent for each panel independently, there may be an inconvenience that the image remains in a portion where no image is present on any panel or at the same position of a plurality of panels. In order to prevent this, the high frequency component of each pixel in the panel may be compared with the component at the same position in the other panel so that only the pixel with the highest high frequency component is left.
Also, the portion where the image is formed between the panels has more high-frequency components than the other panels, but the same level of high-frequency components are generated in two consecutive panels. By leaving, you can make the image appear to be in the middle of the panel due to the visual interpolation effect.

In addition, through the semi-transparent object, since the image with a relatively high frequency component can be seen at the same position on multiple panels, the semi-transparent object can be left by leaving these images as well. Enables natural shooting.
The part without any high-frequency component in any panel is an image outside the shootable distance range, but when transparency processing is performed, a hole is opened in the part of the image, so the image of the panel with the least low-frequency component among them Leave. The amount of the low-frequency component can be examined with a low-pass filter as shown in FIG. The filtering process is the same as described above, and is performed by multiplying and adding the corresponding pixel value and the count value.
Note that these unnecessary video transparency processings may be performed in a stereoscopic video display means described later.

  The imaging device 2 is composed of a single imaging panel 21 and the lens 1 or the imaging device 2 is moved back and forth by placing the imaging device 2 on a linear motor or the like as a means for varying the distance to the object to be photographed. The distance between the panels may be changed, a voltage deformation element may be used as the material of the lens 1, and the lens curvature may be changed by applying a voltage, resulting in changing the focal length f of the lens. By using a plastic material as the material 1 and applying a force in the radial direction of the lens, the curvature of the lens may be changed, or a material whose refractive index changes with voltage such as liquid crystal is used for the lens 1. Thus, the focal length f of the lens may be changed by changing the refractive index of the lens with voltage.

  When there is another object in front of the object to be photographed, as shown in FIG. 7, a part of the rear object is blocked by the front object, and a part of the image is missing. The amount of chipping increases as the lens becomes smaller.However, if it is larger than the amount of chipping when a person looks directly at an object, the image of the part that should be visible is lost and the naturalness is lost. More than binocular distance (about 65 mm) is required.

Next, a method for colorizing a stereoscopic image to be shot will be described.
In general, color images are taken by dividing the area of one pixel of the image pickup device into three parts and attaching color filters of three colors on each surface to obtain a color image. When shooting 3D images with overlapping panels, the same position of each panel is not necessarily the same color, so if you attach a color filter to each panel, the light of the required color will not reach the rear panel. In addition, in the configuration in which one three-color filter is pasted on the forefront, one pixel of the color filter and one pixel of the imaging panel are not aligned with each other in the rear panel, which is inconvenient.

[Embodiment 2]
FIG. 8 is a configuration diagram of a stereoscopic video generation system according to Embodiment 2 of the present invention for solving this problem. In the figure, 1 is a convex lens, 4 is a spectroscope in which two half mirrors 41 are orthogonally crossed, 3 is an imaging device, and one of three color filters 42 is pasted on the surface thereof.
In the color stereoscopic image generation system of this configuration, the light emitted from the photographic object passes through the lens 1 and is split in three directions by the half mirror 41, and each split light has one of the three color filters 42. It passes through and becomes monochromatic light, and is imaged by each of the three imaging devices. However, since each image is monochromatic, there is no need to divide the color filter for each pixel, and no color contradiction occurs between the color filter and the imaging device. .

[Embodiment 3]
Next, a method for stereoscopically displaying different images for each distance obtained by the means will be described.
FIG. 9 is a configuration diagram of a stereoscopic video generation system according to Embodiment 3 of the present invention, and is configured from a transparent stereoscopic video display device 3. FIG. 10 shows a detailed structure of the display device 3, which is composed of a plurality of transparent display panels 31.

FIG. 11 is a structural diagram in the case where a liquid crystal (LCD) is used for the display panel 31, and a driving circuit 311 formed of transparent silicon and a liquid crystal sandwiched between transparent electrodes are placed on a transparent substrate such as glass. The display device is configured by stacking a plurality of the display panels, placing a pair of polarizing plates whose polarization directions are shifted by 90 degrees on the outermost edge, and illuminating from the back surface.
By displaying different object images for each of the shooting distances on the display panels 31 at the respective positions configured as described above, a stereoscopic image having a depth can be obtained. Note that distant objects and images are captured by the imaging panel closer to the lens, so when displaying, the image must be displayed on the farthest display panel as viewed from the viewer. I must. Furthermore, since the photographed image is an inverted image in which the top, bottom, left, and right are inverted, an image with the correct orientation can be reproduced by displaying the image by inverting the top, bottom, left, and right of the image.

From FIG. 4, it can be seen that stereoscopic images captured by an imaging panel with equal intervals are in a correct distance relationship if they are displayed with the same size on a display panel arranged at the same interval. If it is desired to display on a large panel, the interval between the panels may be increased in accordance with the enlargement ratio of the image.
If you look directly at the display device 3, the image can be seen at a depth position that is smaller than the actual distance, but the size of the image is also reduced as the distance is reduced, and the viewing angle, which is the angle at which the object is viewed at the time of shooting, remains unchanged. So you can see a natural 3D image.

  In order to colorize the display image, three color light sources are provided as light sources for illuminating the liquid crystal panel from behind, and the display panel image is switched in synchronism with time-division illumination. Alternatively, as shown in FIG. 12, the display device 3 may be provided for three colors, and each may be irradiated with light sources of different colors, and the resulting image may be synthesized by the half mirror 41. The three-color light sources may be light-emitting diodes of three colors, or may be made by separating the wavelength of white light such as a light bulb with a color filter, as shown in the figure.

FIG. 13 is a structural diagram when a self-luminous element such as an organic Electro-Luminescent (EL) element is used for the display panel 31, and is sandwiched between a drive circuit 311 formed of transparent silicon, a transparent anode and a cathode. The organic EL layer 313 is placed on a transparent substrate such as glass, and a plurality of display panels are stacked to form a stereoscopic video display device.
The feature of this display device is that each panel is self-luminous, so there is no need to illuminate from behind, but each panel has a different number of other panels that transmit light forward, so every panel To achieve the same brightness, the rear panel must have higher brightness.
Further, in order to prevent unnecessary rear light from being transmitted through the front image, the liquid crystal display panel shown in FIG. 11 is used as a liquid crystal shutter and is attached to the rear surface of the display panel to emit light from the display panel. What is necessary is to make the part opaque.

  In order to colorize the display, each pixel of the display panel may be composed of a set of light emitting elements of three colors, or a three-color display device is provided to create a three-color image, as shown in FIG. The three-color image may be synthesized by the half mirror 41 to obtain a color image. In the type in which the light emitting elements of three colors convert white light into three colors with a color filter, or in the type that converts blue light into green or red with a color conversion material, the light emitted from the rear panel is the color of the front panel Therefore, it is better to use a display device with three different colors.

  To increase the sense of depth, the panel spacing can be increased, but to achieve the correct depth ratio, the panel size must be increased at the same time. As another means, a convex lens 1 may be provided on the front surface of the display device 3 as shown in FIG. At this time, if the distance between the lens and the display device is set within the focal length f of the lens, an erecting virtual image with an enlarged depth can be seen behind the display device. At this time, if the distance from the lens to the virtual image is D, the distance from the display panel on which the image is displayed to the lens center is D × f / (D + f), and the magnification of the image is (D + f) / F times.

[Embodiment 4]
When the display device 3 is disposed beyond the focal length f of the convex lens 1, a projection-type stereoscopic video generation system is obtained. FIG. 15 is a configuration diagram of a stereoscopic video generation system according to Embodiment 4 of the present invention. The stereoscopic video display device 3, the front thereof, the convex lens 1 arranged farther than the focal length of the lens, and further arranged in front thereof Screen 5. In FIG. 15, the display device 3 may be a light transmissive display device such as a liquid crystal or a self-luminous display device such as an organic EL element.

  If the focal length of the lens is f and the distance from the lens to the screen is D, the position of the image focused on the screen in the original display device 3 is the distance D × f / (D -F), the display device may be arranged so that the corresponding display panel comes to this distance. Since it is necessary to change the position of the screen in accordance with the depth position of the displayed image, a plurality of translucent screens are stacked in order on each position, or a directional reflection screen as shown in FIG. 16 is used. .

  In FIG. 16, reference numeral 51 denotes a minute spherical lens, which is a reflecting mirror whose half surface is a mirror surface. The focal length f of such a spherical lens is f = r / 2 × n / (n−1), where r is the radius of the sphere and n is the refractive index of the lens material, so n = 2. F = r, and the focal position matches the lens surface. Incident light is focused on the bottom surface of the lens, reflected there, and exits in the same direction as the incident light. Since the light reflected in this manner appears to have come out of the focal position when there is no screen, it is possible to see the real images formed at the respective depth positions. Since normal glass has n = 1.5, f = 1.5r. In this case, as shown in FIG. 17, a cylindrical lens type reflecting mirror having a thickness of 3r may be used, and the curvature of the front surface may be 1 / r and the curvature of the back surface may be 1 / 2r.

[Embodiment 5]
FIG. 19 is a configuration diagram of a stereoscopic video generation system according to Embodiment 5 of the present invention. In FIG. 19, 61 is a high frequency component extracting means, and 62 is a luminance changing means. In the configuration shown in the figure, video signals output from each imaging panel in the imaging device 2 are input to the high frequency component extraction means 61, and high frequency components are extracted by a Laplacian filter as shown in FIG. The Laplacian filter outputs a positive and negative large-amplitude signal pair as the high-frequency component of the input signal is stronger. Therefore, the absolute value may be used as the amount of the high-frequency component, that is, the degree of focus.

  However, even in an in-focus image, there are few high-frequency components in areas where there are originally fine images, and there is a risk of being judged as a blurred area.Therefore, a morphological filter, median filter, etc. are scattered in the focused area. It is better to fill the region with less high frequency components. At this time, even in the video region that should originally be blurred, the high-frequency component is extracted in the portion where the contrast of the subject is strong, and the region is enlarged. In order to remove this, the morphological filter is then applied in the reverse order. It is preferable to remove the high-frequency components scattered in the blurred region by applying or applying the median filter again. Even in this way, the once connected high frequency component region is not narrowed, and the amount of high frequency component can be extracted stably.

  The extracted high-frequency component amount is compared between images from adjacent imaging panels, the high-frequency component amount in the region with the highest high-frequency component amount is associated with 1, and the high-frequency component amount at the same position on the other panel is 0. , The high-frequency component amount is normalized from 0 to 1 between the two panels. That is, the normalized high-frequency component amount X in the focused area on the panel 1 is 1, and the normalized high-frequency component amount in the corresponding panel 2 is 1−X = 0. In addition, the normalized high-frequency component amounts in the regions that are blurred to the same extent on both panels are X = 0.5 and 1−X = 0.5. This value X is the position where the image originally focuses. Is equal to the value normalized by the distance between panels.

Next, the luminance changing unit 62 changes the luminance Y of the blurred image from the normalized component amount X, and obtains new luminances Y1 and Y2 as follows. Here, the luminance values of the two panels are almost equal because they are blurred images of the same subject, and are Y.
When the display panel controls the amount of light transmitted from the light source, such as a liquid crystal display panel, the original luminance value Y changes between 0 and Z (for example, 255).
Y1 = Z ^1-X × Y ^{X
Y2 = Z X × Y 1-} X
And change. When two panels with this luminance value are overlapped and illuminated from the back, the luminance of the output video is the product of both.
Y1 × Y2 = Z × Y
It becomes.

  This is the same as viewing the brightest panel (Y = Z: transparent) and the panel with the original luminance value Y, and an image with the original luminance value is obtained. Due to the interpolation action, the two panels are biased toward the higher contrast. That is, the smaller the high frequency component, the closer the value of X approaches 1 to 0, Y1 becomes brighter than Y2, and the contrast of Y1 falls below Y2, so that the image is localized at a position closer to the Y2 panel. Conversely, as the high frequency component amount X is closer to 1, the image is localized at a position closer to the panel side of Y1.

If the display panel is self-luminous, such as an organic EL panel, the original luminance value Y is
Y1 = X × Y
Y2 = (1-X) × Y
And change. If you look at two panels with this brightness value, the brightness of the output video will be the sum of both.
Y1 + Y2 = Y
Thus, an image having the original luminance value is obtained, but its position is biased toward the higher contrast between the two panels due to the interpolation of the eyes. That is, as the high frequency component is smaller, the value of X approaches 1 to 0, Y1 is darker than Y2, and the contrast of Y1 is lower than Y2. Therefore, the image is localized at a position closer to the Y2 panel. Conversely, as the high frequency component amount X is closer to 1, the image is localized at a position closer to the panel side of Y1.

  As described above, the original position of the subject can be determined from the amount of blur of the video, and the luminance distribution corresponding to the subject is performed on the display panel, so that the video between the panels including the position can be reproduced.

  The 3D image generation system according to the present invention is useful as a system for generating natural 3D images because it stores and displays the relationship between the size of an object existing in a three-dimensional space in the real world and the distance to the object. In particular, it eliminates all the disadvantages of conventional stereoscopic video systems that use only binocular parallax, and does not require spectacles, but does not cause video discontinuity or depth reversal due to the monitoring position, and focus adjustment. Therefore, it is useful as a 3D image generation system that does not cause discomfort or fatigue even when viewed for a long time.

The block diagram of the imaging means of the three-dimensional video generation system in Embodiment 1 of this invention. The block diagram of a three-dimensional video imaging device. The block diagram of a three-dimensional image pick-up panel. The figure which shows the imaging object distance and blurring amount for every imaging panel position. The block diagram of a high frequency component detection filter. The block diagram of a low frequency component detection filter. Illustration of occlusion. The block diagram of the color imaging means of the three-dimensional video production | generation system in Embodiment 2 of this invention. The block diagram of the image | video display means of the stereo image production | generation system in Embodiment 3 of this invention. The block diagram of a three-dimensional video display device. 1 is a configuration diagram of a light transmission type stereoscopic image display panel. FIG. The block diagram of the color-ized video display means of a three-dimensional video production | generation system. 1 is a configuration diagram of a self-luminous stereoscopic image display panel. FIG. 1 is a configuration diagram of a stereoscopic video generation system that displays an enlarged video. The block diagram of the image | video display means of the three-dimensional image | video production | generation system in Embodiment 4 of this invention. The block diagram of a directional reflective screen. The structural diagram of a cylindrical lens type reflecting mirror. The block diagram of the conventional binocular stereo camera. The block diagram of the three-dimensional video generation system in Embodiment 5 of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Convex lens, 2 ... Imaging device, 21 ... Imaging panel, 211 ... Transparent electrode, 3 ... Display device, 31 ... Display panel, 311 ... Drive circuit, 312 ··· Polarizing plate, 313 ... Organic EL layer, 4 ... Spectroscope, 41 ... Half mirror, 42 ... Color filter, 5 ... Screen, 51 ... Spherical mirror, 61 ... .High frequency component detection means, 62... Brightness change means

Claims (52)

A system for generating stereoscopic images,
It has a stereoscopic video imaging means composed of a lens that forms a real image of an object and an integrated imaging device arranged at the real image position,
The stereoscopic image capturing unit obtains a plurality of images according to the distance to the imaged object by changing the distance from the lens to the imaging position of the imaging device according to the distance to the object, and does not form an image. A 3D image generation system characterized by making the image of the part transparent.   The three-dimensional image generation system according to claim 1, wherein a lens having a large aperture is used to transparentize an image of a portion where no image is formed.   The imaging device has a configuration in which a plurality of transparent imaging panels are superposed so that the distance from the lens to the imaging position is different, and multiple images corresponding to the distance to the imaged object are obtained. The stereoscopic video generation system according to claim 1, wherein:   The stereoscopic image generation system according to claim 3, wherein the imaging panels are arranged at equal intervals so that a maximum amount of blur of an image formed on any of the imaging panels is a predetermined value or less. .   The stereoscopic image generation system according to claim 3, wherein the imaging device is configured to change a range of a distance that can be imaged by changing an interval between imaging panels.   The stereoscopic image generation system according to claim 3, wherein the imaging panel is a transparent imaging panel by having a configuration in which optical sensor elements are integrated on a transparent substrate.   The stereoscopic image according to claim 6, wherein the photosensor element has a higher sensitivity as it is farther from the lens, so that an image with the same brightness can be obtained from any of the superimposed imaging panels. Generation system.   The stereoscopic image generation system according to claim 6, wherein the optical sensor element is configured such that light reaches the rear sensor element by making the electrode transparent.   The stereoscopic image generation system according to claim 1, wherein the imaging device is configured to be able to capture a real image by disposing the imaging device at a distance greater than a focal length of the lens.   The imaging device is arranged within a distance D × f / (D−f) from the lens, where f is the focal length of the lens and D is the distance to the most recent object to be imaged. The three-dimensional video generation system according to claim 9, wherein the video is obtained.   The stereoscopic video imaging means treats the blurred part by bringing the signal value of this part close to the maximum value by taking the part not including the high-frequency component in the video signal output from the imaging device as the part not imaged. 2. The stereoscopic image generation system according to claim 1, wherein the stereoscopic image generation system is made transparent and removed.   The stereoscopic video imaging means compares the video signal output from the imaging panel with the video signal output from the same position in another imaging panel, and determines the video signal containing the most high-frequency components as the signal with the smallest blur amount. 12. The stereoscopic video generation system according to claim 11, wherein unnecessary video signals are removed so as to remain.   The stereoscopic video imaging means outputs high-frequency components of the video signal output from the imaging panel at substantially the same level as the video signal output from the same position of the other imaging panel, and is output from the same position in the other imaging panels. If the image signal contains more high-frequency components than the image signal, the image of the object is considered to be an image formed in the middle of the two imaging panels, or the overlapping part of the semitransparent front image and the rear image. The 3D image generation system according to claim 11, wherein the 3D image generation system is left.   When a video signal including a high frequency component is not output from any of the imaging panels, the stereoscopic video imaging means treats the video signal with the least low frequency component as an image outside the shootable distance range. 12. The three-dimensional video generation system according to claim 11, wherein a video signal output from the same position of the other imaging panel is left transparent and removed.   The stereoscopic video generation system according to claim 11, wherein the stereoscopic video imaging means improves blurring of a video image formed between the imaging panels by passing the remaining video signal through a high-pass filter.   2. The stereoscopic image capturing means is characterized in that a plurality of images are obtained by using a variable focus lens that changes its focal length in accordance with an externally applied voltage or physical force. 3D image generation system.   The three-dimensional video generation system according to claim 1, wherein the lens and the imaging device have a movable mechanism that changes a distance between the lens and the imaging device so as to obtain an image according to the distance to the subject.   The three-dimensional image generation system according to claim 1, wherein a lens having a large diameter is used to realize a difference in occlusion due to binocular parallax and to approximate a more natural image.   The three-dimensional image pickup means has a spectroscope that splits light in three directions between the lens and the image pickup device, and the image pickup device that receives each of the split light has three colors different from those of other image pickup devices on the front surface. The stereoscopic video generation system according to claim 1, wherein a color stereoscopic video image can be taken by adopting a structure having any one of the color filters. A system for generating stereoscopic images,
Having a 3D image display means including an image display device for displaying an image of an object,
The three-dimensional video display means is a three-dimensional video having a natural depth by varying the distance to the video display position in accordance with the ratio between the size of the video to be displayed and the distance to the position where the video should originally be. A three-dimensional video generation system characterized by displaying video.   The video display device has a configuration in which a plurality of transparent video display panels are overlapped to change the distance to the position of the display video and display a stereoscopic video with depth. Item 20. The stereoscopic image generation system according to Item 20.   The video display panel is configured to be arranged at equal intervals, so that the amount of blurring of the displayed video is less than a certain value regardless of the depth, and even 3D video is displayed. The stereoscopic video generation system according to claim 21.   The video displayed on the video display device is displayed upside down at the top and bottom, left and right, front and back, and inverted to display the stereoscopic image with the depth reversed at the normal position. The three-dimensional video generation system according to claim 21, wherein:   The stereoscopic video generation system according to claim 21, wherein the video display device is configured to change a depth range of a stereoscopic video that can be displayed by changing an interval between video display panels.   The three-dimensional video generation system according to claim 21, wherein the video display panel has a configuration in which a transmission light amount variable element is integrated on a transparent substrate, and the video is displayed by illuminating it from behind.   The three-dimensional video display means has a light source that sequentially emits light of each of the three primary colors at different timings and irradiates the video display device from the back side, and the video display device synchronizes with the luminescent color. 26. The stereoscopic video generation system according to claim 25, wherein a color stereoscopic video is displayed by displaying the video of   The stereoscopic video display means has a light source and a video display device for each of the three primary colors, and displays a color stereoscopic video by having a light synthesizer that combines the three colors of video output from each video display device into a single color video. 26. The 3D image generation system according to claim 25, wherein   23. The stereoscopic video generation system according to claim 21, wherein the video display panel has a configuration in which transparent self-light-emitting elements are integrated so that a stereoscopic video that is not affected by the video light on the rear panel can be displayed.   The video display panel is characterized in that an optical shutter is provided on the back side of the light emitting element so that the light from the rear is blocked during light emission, thereby enabling to display a stereoscopic video that is not affected by the video light on the back panel. The three-dimensional video generation system according to claim 28.   29. The stereoscopic video according to claim 28, wherein the video display panel has a brightness that is uniform regardless of the depth by correcting the attenuation of the amount of light in the front panel by increasing the brightness of the rear panel. Generation system.   21. The stereoscopic image display device according to claim 20, wherein the stereoscopic image display means has a configuration in which a convex lens is provided on the front surface of the image display device, thereby apparently increasing the depth of the object image displayed on the image display device. Video generation system.   32. The three-dimensional video generation system according to claim 31, wherein the video display device is arranged so as to observe an erect virtual image by being arranged within a focal distance of the lens from the lens center.   The stereoscopic image display means makes a portion that does not contain a high-frequency component in a video signal to be displayed as a portion that has not been imaged and makes the signal of this portion close to the maximum value, thereby transparently removing the blurred portion. 21. The 3D image generation system according to claim 20, wherein:   The stereoscopic video display means compares the amount of blur of the video signal to be displayed with the video signal displayed at the same position in another display panel, and displays the video signal containing the highest frequency component as the minimum amount of blur. 34. The three-dimensional image generation system according to claim 33, wherein:   The stereoscopic video display means that the high-frequency component of the video signal output to the display panel is displayed at the same position in the other display panels at the same level as the video signal displayed at the same position in the other display panels. If the image of the object contains more high-frequency components than the image signal, the image of the object is considered to be an image formed in the middle of both display panels, or it is regarded as an overlap between the semitransparent front image and the rear image, 34. The three-dimensional video generation system according to claim 33, wherein the three-dimensional video generation system is displayed.   When the video signal displayed on any of the display panels does not contain a high frequency component, the stereoscopic video display means displays the video signal with the lowest low frequency component as the video outside the shootable distance range. 34. The stereoscopic video generation system according to claim 33, wherein the video signal displayed at the same position of the other display panel is left transparent and removed.   34. The stereoscopic video generation system according to claim 33, wherein the stereoscopic video display means improves blurring of an image formed between display panels by passing the display video signal through a high-pass filter.   The stereoscopic image display means displays the projection type stereoscopic image by disposing the image display device at a distance longer than the focal distance of the lens from the center of the lens and providing a screen in front of the lens. 32. The 3D image generation system according to claim 31, wherein:   The screen is a directional reflective screen that reflects incident light in the incident direction, so that a real image formed at an arbitrary distance can be seen regardless of the installation position of the screen. 38. The stereoscopic video generation system according to 38.   40. The three-dimensional image generation system according to claim 39, wherein the directional reflection screen has a configuration in which minute spherical mirrors with a refractive index of 2 are spread to reflect incident light in the incident direction.   The directional reflection screen has a configuration in which minute bullet-shaped cylindrical mirrors with a refractive index of 1.5 are spread, the surface curvature is 1 / r, the back surface curvature is 1 / 2r, and the distance between both surfaces is 3r. 40. The three-dimensional image generation system according to claim 39, wherein incident light rays are reflected in the incident direction.   The stereoscopic video imaging means changes the luminance value of the video according to the amount of high-frequency components included in the video signal output from each imaging panel constituting the imaging device, so that the display panel and the display panel constituting the display device are changed. 3. The stereoscopic video generation system according to claim 1, wherein an apparent position of the display video is localized in between.   The stereoscopic video imaging means distributes the luminance value of the video between adjacent panels according to the amount of high-frequency components included in the video signal output from the imaging panel that constitutes the imaging device. 2. The stereoscopic video generation system according to claim 1, wherein the display video is localized between a display panel constituting the display device and the display panel in accordance with the position of the video image formed on the display device. In the stereoscopic video imaging means, values obtained by normalizing the amount of high-frequency components included in video signals output from adjacent imaging panels constituting the imaging device between 0 and 1 are X and (1−X). When the luminance value Y of the video changes in the range from 0 to Z, the luminance value Y is changed to Z ^{1-1 / X} × Y ^{1 / X} and Z ^{1 / X} × Y ^{1-1 / X} to make it transparent By displaying the images on the adjacent display panels constituting the transmitted light amount control type display device, the display image is displayed at a position between the display panels in proportion to the original position of the photographed object. The stereoscopic image generation system according to claim 1, wherein localization is performed.   The stereoscopic video imaging means obtains the amplitude at the zero cross point of the value obtained by passing the video signal output from each imaging panel constituting the imaging device through the Laplacian filter, and smoothes it through at least one of the morphological filter and the median filter. By changing the brightness of the video according to the amount of high-frequency components and displaying it on each display panel that constitutes the display device, the position of the displayed video is smooth at a position proportional to the original position of the photographed object. The stereoscopic image generation system according to claim 1, wherein localization is performed.   The stereoscopic video imaging means changes the luminance of the video according to the amount of high frequency components included in video signals output from a plurality of imaging panels constituting the imaging device, and performs composite display of the video image formed in the middle of the panel 2. The stereoscopic image generation system according to claim 1, wherein the position of the synthesized image is stably localized by giving priority to the high-frequency component amount of the front panel.   The stereoscopic image display means localizes the apparent position of the display image between the display panel constituting the display device by changing the luminance of the image according to the amount of high frequency components contained in the displayed video signal. 21. The three-dimensional video generation system according to claim 20, wherein   The stereoscopic image display means distributes the luminance of the image between adjacent display panels constituting the display device in accordance with the amount of high-frequency component included in the displayed video signal, so that the original position of the captured object is 21. The three-dimensional video generation system according to claim 20, wherein the display video is localized between the display panels according to the above. In the stereoscopic image display means, the value obtained by normalizing the amount of high-frequency components included in the image signal displayed on the adjacent display panel constituting the transmitted light amount control type display device between 0 and 1 is X and (1− X), and when the luminance value Y of the original video changes in the range of 0 to Z, the luminance value Y is changed to Z ^{1-1 / X} × Y ^{1 / X} and Z ^{1 / X} × Y ^{1/1 / X.} 21. The stereoscopic image generation according to claim 20, wherein the position of the display image is localized at a position proportional to the original position of the photographed object by making each of them transparent and displaying them. system.   The stereoscopic image display means has a value obtained by normalizing a high-frequency component amount included in an image signal displayed on an adjacent display panel constituting a self-luminous display device between 0 and 1, and X and (1-X ), The luminance value Y of the original video is changed to X × Y and (1−X) × Y, respectively, and is displayed at a position proportional to the original position of the photographed object. 21. The three-dimensional video generation system according to claim 20, wherein the position of the video is localized.   The stereoscopic video display means obtains the amplitude at the zero cross point of the value obtained by passing the video signal to be displayed on each display panel constituting the display device through the Laplacian filter, and smoothes it through at least one of the morphological filter and the median filter. 21. The position of the display image is smoothly localized at a position proportional to the original position of the photographed object by changing the luminance of the image according to the amount of the high-frequency component. The described stereoscopic image generation system. The stereoscopic video display means changes the luminance of the video according to the amount of high frequency components included in the video signal displayed on the plurality of display panels constituting the display device, and performs composite display of the video image formed in the middle of the panel. 21. The three-dimensional image generation system according to claim 20, wherein the position of the synthesized image is stably positioned by giving priority to the amount of high-frequency components in the front panel.

Images