JPH0553255B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a method for photographing a three-dimensional video of a subject having a fixed external shape that does not change at least during photographing, such as a Buddhist statue, using a movie camera or a television camera. In particular, it relates to a method that can arbitrarily adjust the stereoscopic effect.

[Conventional technology]

Conventionally, in order to capture a stereoscopic image with binocular parallax as a moving image, a left camera 1 and a right camera 2 are placed in parallel with a certain distance d apart, as shown in the plan view of FIG. Then, while moving both cameras 1 and 2 closer to the subject S (track-up) or moving them away from the subject S (track-back), the video seen with the left eye and the video seen with the right eye of the subject S are simultaneously taken to create a 3D video. I was filming. Both cameras 1 and 2
Distance d of the straight line connecting
This is the distance connecting the object-side principal points of the objective lenses L1 and L2, but it does not necessarily have to be the distance between the positions of the principal points. ) is called the baseline length d, and this straight line is called the baseline l. Conventionally, since the distance between the human eyes is constant, the baseline length d is held constant and the line is bisected perpendicularly to the baseline l. The subject S was positioned on a line, and the cameras 1 and 2 were moved closer to the subject S and moved away from it while taking pictures. To adjust the viewing angle of each camera at this time, either change the rotation angle so that cameras 1 and 2 themselves face the subject S, or attach a mirror or prism in front of the objective lens of the camera and change the angle. I was going by.

[Problem to be solved by the invention]

However, such conventional stereoscopic video shooting methods have the following problems, and therefore it has been difficult to obtain easy-to-see stereoscopic videos with natural stereoscopic effect.

That is, first, both cameras 1 and 2 are set to the subject S.
As you move closer to , the objective lens L1 of camera 1,
L2 will come into physical contact with each other and the convergence angle α of cameras 1 and 2 will not be able to exceed a certain level, so when the film with binocular parallax is projected as a stereoscopic image, the image of the subject S will be Because it appears to stand out in front of the screen, it becomes extremely difficult to observe 3D video, and in extreme cases, it becomes difficult to recognize it. To explain this point a little further, Figure 8a
As shown in , the lines of sight of both cameras 1 and 2 during shooting (usually the optical axis, but if the point of interest is off the center of the screen, the line of sight connecting the point of interest and the principal point of the objective lens) When the point where the two intersect (hereinafter referred to as the cross point P) is located after the subject S
(A), when located in or on the surface of the subject S (B),
There is a case (C) where the camera is located in front of the subject S. Corresponding to each case, when the photographed film is projected onto the screen 3 for stereoscopic observation and the observer M observes it, the image of the subject S is
S' appears to be located in front of the screen 3 in case A, on the screen 3 in case B, and on the other side of screen 3 in case C. When observing, it is easy to see the image S' if it is located on or near the screen 3.
By the way, as mentioned above, the two cameras 1 and 2
When moving closer to the subject S, the objective lenses L1 and L2 of the cameras 1 and 2 collide, and the convergence angle α of the cameras 1 and 2 cannot exceed a certain level, so
If the object is brought closer than that, a three-dimensional image must be taken in the above state A, so when the photographed film is projected, it will be in the state A in Fig. 8b, i.e.
Since the image S' of the subject S appears to stand out in front of the screen, it is very difficult to observe the stereoscopic video, and in extreme cases, it becomes difficult to recognize it. In particular, small objects must be photographed at close range, but it has been impossible to obtain observable stereoscopic images of them because of collisions between lenses.

In addition, the second problem is that the two cameras 1,
In a stereoscopic image obtained by photographing the subject S close-up while keeping the baseline length d of 2 constant, the stereoscopic effect appears to be unnaturally emphasized. For example, if you look at an image taken very close to a sphere, it looks like an ellipsoid with its long axis extending back and forth. In order to prevent this, it is possible to make the baseline length d, which is the distance between the two cameras 1 and 2, variable and shorten the baseline length d during close-up photography. , 2 cannot be set below the distance at which the objective lenses L1 and L2 of 2 are in contact with each other.

The present invention has been made in view of the above circumstances, and its purpose is to use a movie camera or a television camera to shoot a three-dimensional video of a subject having a fixed external shape that does not change during filming, such as a Buddhist statue. To provide a method capable of arbitrarily adjusting the stereoscopic effect and photographing a stereoscopic moving image that is easy to observe when approaching or moving away from a subject and photographing the stereoscopic moving image.

[Means to solve the problem]

A method for photographing a three-dimensional video of a subject having a fixed external shape according to the present invention, which achieves the above object, is a method for photographing a three-dimensional video of a subject whose external shape does not change during at least the shooting, using a single movie camera or television camera. In order to position the image of a specific point on the shooting screen at a certain position, display the image of the subject on the viewfinder or monitor screen of a movie camera or television camera, select a specific point on the subject, and then determine the position of that point. The camera is displayed so that it can be identified during shooting, and the center line is a straight line connecting the camera at that position and a specific point on the subject, and the camera is moved in one direction perpendicular to the center line by a distance determined for each shooting position. moving the subject, adjusting the line of sight of the camera so that the image of the specific point on the subject matches the displayed position, and photographing one frame on the left and right of the subject;
Also, move the camera the same distance in the opposite direction from the center line, adjust the line of sight of the camera so that the image of a specific point on the subject matches the displayed position, and shoot the other frame on the left and right. This method is characterized by photographing a group of stereoscopic video frames by performing the following steps at positions where the distance of the camera to the subject differs slightly.

In this case, if the distance determined for each shooting position of the camera is set to a distance that decreases linearly as the distance from the subject decreases, the 3D effect will remain unchanged even as the distance from the subject decreases. You can get the video. On the other hand, if the distance decreases along a curve in which the rate of decrease increases as the distance from the subject decreases, a 3D video with a natural 3D effect can be obtained.

In addition, adjusting the camera's line of sight so that the image of a specific point on the subject matches the displayed position,
You can measure the distance between a specific point on the subject and the camera, calculate the direction of line of sight from that measurement and the distance the camera has moved relative to the center line, and use the camera's viewfinder or monitor to calculate the direction of the line of sight. You can also do this while observing the image of the subject on the screen.

[Effect]

In the method of photographing a stereoscopic video of a subject having a fixed external shape according to the present invention, a single movie camera or television camera is used, and the movie camera or television camera is Display the image of the subject on the viewfinder or monitor screen, select a specific point on the subject, display the position of that point so that it can be determined during shooting, and connect the camera at that position and the specific point on the subject. Using the connecting straight line as the center line, move the camera in a direction perpendicular to the center line an equal distance to the left or right by a distance determined for each shooting position, and move the camera so that the image of the specific point on the subject matches the displayed position. Since the left and right frames of the subject are photographed by adjusting the line of sight, the line of sight always intersects at one specific point on the subject, and the projected 3D image is always located near the screen. You can get an easy-to-view 3D video. Furthermore, since the baseline length for shooting a stereoscopic video can be determined for each shooting position, the stereoscopic effect to be obtained can be arbitrarily selected, and not only a natural three-dimensional effect but also a supernatural three-dimensional effect can be obtained. Furthermore, since only a single movie camera or television camera is used, it is possible to get very close to the subject and shoot a 3D video with a natural 3D effect. Furthermore, it is also possible to capture stereoscopic videos of minute objects.

〔Example〕

First, let us consider a case where the human eye views a three-dimensional object. As shown in Figure 6, when looking at an object located at infinity, the lines of sight of both eyes are approximately parallel, the cross point where the line of sight intersects is located at infinity, and the focus is also set at infinity. (Figure a).
When looking at an object at a normal distance, the eye is aligned so that the cross point is on one of the points on the object, and the focus is also set on the cross point accordingly (Figure b). Furthermore, when looking at an object that is very close, the eyes are shifted further so that the cross point is on some point on the object, and the focus is also on the cross point (Figure c).

Therefore, when shooting 3D videos with a movie camera, TV camera, etc., it is desirable to have the camera's line of sight intersect at some point on the subject in order to capture images with a natural 3D effect. , and as explained in relation to Figure 8, in order to prevent the projected 3D image from rising off the screen and becoming difficult to see, it is necessary to set the cross point at some point on the subject. be.

By the way, as mentioned above, when trying to shoot a 3D video with two cameras, a situation arises in which it becomes difficult to position the cross point over the subject, and
There are cases where the three-dimensional effect is too emphasized. However, if the external shape of the subject changes during shooting, such as a human being, it is necessary to simultaneously capture images with parallax using two cameras. In this case, it is not necessarily necessary to take pictures with two cameras at the same time. Therefore, in the present invention, images with left and right parallax are sequentially photographed using one camera. This point is the first feature of the present invention. And when taking images from the left or the right, the cross point is set at a predetermined point on the subject, and the baseline length can be changed depending on the subject distance. be.

This point will be further explained with reference to FIG. As shown in Figure a, the angle of one camera 10 with respect to the subject S is determined, the line connecting the subject S and the camera 10 is set as the center line t, and the subject S to be taken as a reference is set on the viewfinder of the camera 10 or the monitor screen F. The upper cross point P is determined, and the position of the determined cross point P is indicated or stored in the screen F. Then, the distance z between the subject S and the camera 10 is measured. Although the focus signal of the camera can be used for this measurement, it is not necessarily limited thereto, and other means may also be used. Then, as shown in Figure b, the camera 10 is moved perpendicularly to the center line t to the left or right (left in the figure) by half of the baseline length d suitable for that position of the camera 10, And angle β = arctan determined from the relationship of a right triangle with two sides of length z and d/2
The line of sight of the camera 10 is rotated by (d/2z) toward the cross point P. When rotated in this manner, the cross point image indicated or stored on the viewfinder or monitor screen F matches the image of the cross point seen visually. Note that this rotation of the line of sight of the camera 10 may be performed automatically according to the above formula β=arctan(d/2z), or while looking at the viewfinder or monitor screen F, so that the image of the cross point you actually see matches the position of the cross point P that you have instructed or memorized.
The angle may be adjusted. and,
An image of the subject S at that position is taken. Then,
Next, as shown in FIG. 1c, move the camera 10 to the opposite side with respect to the center line t by half of the base line length d, and similarly change the viewing angle to β with respect to the center line t.
Adjust it only inward and take the image. The two images taken in Figures 1 b and c above are paired to form a stereoscopic image reproduction film with binocular parallax, and this pair is projected onto a screen using, for example, polarized glasses. When viewed, an easy-to-see image S' of the subject S is reproduced on or near the screen S, as shown in FIGS. 8B and 8B.

Next, for example, when capturing a video in which the camera 10 approaches the subject S, the camera 10 is brought close to the subject S, and the distance Δz of the approach is measured, for example, by measuring the moving distance of the movable table on which the camera 10 is placed. By measuring and subtracting the distance Δz from the previous distance z, the distance between the subject S and the camera 1 is calculated.
The distance between 0 and 0 is determined, and a pair of left and right images are photographed in the same manner as in b and c in FIG. At this time, without moving the position of the cross point P that was first specified or memorized on the viewfinder or monitor screen F, calculate the angle by formula so that the position matches the image of the cross point that you are actually seeing. Adjust accordingly or manually.

In this way, the cameras 10 gradually approach each other.
A pair of images for a stereoscopic video is captured using the method, but whether the baseline length d when approaching the subject S is fixed or variable depends on how the stereoscopic effect of the resulting stereoscopic video is selected. Dependent. As described above, in a stereoscopic image obtained by photographing the subject S close-up while keeping the baseline length d constant, the stereoscopic effect appears to be unnaturally emphasized. If this point is not a problem, as shown in FIG.
(It doesn't have to be constant, it changes depending on the close-up) While gradually approaching the subject S and aligning the line of sight with the point P on the screen F that was originally decided as described above, Go and take pictures of the pair. The shooting order is not necessarily →→
It doesn't have to be →→→, but →→→
→→, →→→→→, etc., and the order of the frames to be projected is selected at the editing stage depending on whether the camera 10 is projecting in the direction toward which it approaches or in the direction in which it moves away. , Furthermore, the above-mentioned photographing order is selected from the viewpoint of ease of photographing. In the case of Fig. 2, as mentioned above, when the image of the subject S is taken in an extremely close-up, the three-dimensional effect is exaggerated as if it were extended back and forth, but as shown in the trajectory shown in Fig. 3, , when the camera 10 is brought close to the subject S along the hypotenuse of an isosceles triangle with the cross point P as the apex, the base as d, and the height as z, the line of sight of the camera 10 does not change on each hypotenuse, and the three-dimensional The feeling does not change even when it is close up and remains as it was at the beginning. Such a three-dimensional effect is not the same as the natural three-dimensional effect produced by a normal close-up of both eyes. Therefore, the expression of a natural three-dimensional effect is expressed by the above-mentioned figures 2 and 3, as shown by k in Fig. 4.
Although it is considered to be on the trajectory between the figures, this is because the three-dimensional effect differs between individuals, the three-dimensional effect differs depending on the subject, and the three-dimensional effect differs depending on the focal length of the objective lens of the camera 10. , cannot be determined uniquely. Therefore, the photographer selects the trajectory k that gives the most natural three-dimensional effect while changing the baseline length between the left and right camera positions depending on the distance between the camera 10 and the subject S. Moreover, by moving the camera 10 along other trajectories, for example, along trajectories located inside the isosceles triangle in FIG. 3, it is also possible to express a three-dimensional image that does not exist naturally. In this case, in order to create a more natural three-dimensional effect when zooming in close-up, the rate of decrease in the baseline length should be greater when using a wide-angle lens than when shooting using a standard lens. Conversely, if a telephoto lens is used, the rate of decrease needs to be done more slowly.

Now, in order to realize the method of shooting a stereoscopic video using one camera 10 as described above, for example, the fifth
The equipment shown in the diagram is required. In the figure a, the objective lens L of the camera 10 must be equipped with a focus adjustment mechanism 11. As described above, it is desirable that a distance signal from the focus adjustment signal to the cross point can be obtained. Furthermore, it is necessary to have a horizontal angle adjustment mechanism 12 that can set the horizontal rotation angle of the camera. It is also necessary to include a horizontal movement mechanism 14 for moving the same distance left and right with respect to the central axis t. Preferably, the camera 10 also includes a vertical angle adjustment mechanism 13 that adjusts the vertical angle of the line of sight of the camera 10. A camera 10 equipped with a focus adjustment mechanism 11, a horizontal angle adjustment mechanism 12, a horizontal movement mechanism 14, and a vertical angle adjustment mechanism 13 as described above.
is installed on the mounting table 17 at the height adjustment position 15 in b of the figure. The height of the mounting table 17 is adjusted by moving along the vertical rails 16. The height adjustment device 15 in FIG. b is placed on the moving table 20 of the camera moving mechanism 18 shown in FIG. 10 may be placed. The moving table 20 moves back and forth along the rail 19,
The rails 19 are arranged parallel to the centerline t (FIG. 1). The moving distance of the movable table 20 is calculated from the rotation speed of its wheels 21. The subject S (Fig. 1) is placed on the mounting table 23 of the subject mounting device 22 shown in d of the figure, but in order to rotate the subject S around the vertical axis during photographing, the subject mounting device 22 has a A rotation adjustment device 24 is provided.
Note that when photographing a stereoscopic image that wraps behind the subject S by rotating it during photographing, it is necessary to adjust the rotation axis so that it passes through the cross point P. The above focus adjustment mechanism 11 is preferably an autofocus mechanism, and also includes a horizontal angle adjustment mechanism 12, a vertical angle adjustment mechanism 13, a horizontal movement mechanism 14, a height adjustment device 15, and a camera movement mechanism 1.
8. It is desirable that the rotation adjustment device 24 be able to automatically adjust the target parameter based on a signal from the CPU, and it is also desirable that the rotation adjustment device 24 be able to adjust it manually. moreover,
Although not shown in the drawings, if the camera 10 is a movie camera, it is desirable that the indicator can be movably pointed to any point within the viewfinder's field of view, and if it is a television camera, it is also possible to move the index to any point within the viewfinder's field of view. The indicator can be movably indicated at any point on the screen, or part or all of the initial screen for determining the cross point can be displayed overlappingly during subsequent shooting within the viewfinder field of view or on the monitor screen. is desirable.

In order to take a stereoscopic video of a subject having a fixed external shape using such equipment, the subject S is set on the mounting table 23 after adjusting as described above.
Next, the height adjustment device 15 is placed on the moving table 20, and the camera 10 is set on the mounting table 17.
Then, the horizontal angle adjustment mechanism 12 and the vertical angle adjustment mechanism 1
3. Adjust the horizontal movement mechanism 14, height adjustment device 15, camera movement mechanism 18, and rotation adjustment device 24,
As shown in Figure 1a, the center line t is determined and the distance z
and determine the position of the cross point P within the viewfinder field of view or on the monitor screen.
Next, the camera 10 is moved by the horizontal movement mechanism 14 by a distance of half the base line length determined at that position to the left and right of the center line t, and the actual screen is displayed on the viewfinder or monitor so that the line of sight passes through the cross point P. The horizontal angle adjustment mechanism 12 can be adjusted manually while comparing the position of the cross point P determined at the beginning, or the horizontal angle adjustment mechanism 12 can be adjusted automatically according to β=arctan (d/2z). do. In this way, left and right stereo pair images at a specific distance are photographed, and then the camera movement mechanism 1
8 to move the camera forward or backward with respect to the subject S, and similarly capture left and right stereo pair images. In this case, the distance moved to the left and right of the center line t (half of the base line length) is , is the distance along one of the trajectories shown in Figures 2 to 3, but this trajectory must be set in the CPU in advance, or
Set the settings step by step while shooting. Note that the photographing order is as explained in connection with Fig. 2.
Various modifications are possible, and it is not necessarily necessary to perform left and right imaging successively.

In the above explanation, the method of photographing a 3D video of a subject having a fixed external shape according to the present invention has been explained based on the assumption that the subject is of a normal size. It is possible to create 3D videos of tiny objects by combining a single microscope and camera.

〔Effect of the invention〕

In the method of photographing a stereoscopic video of a subject having a fixed external shape according to the present invention, a single movie camera or television camera is used, and the movie camera or television camera is Display the image of the subject on the viewfinder or monitor screen, select a specific point on the subject, display the position of that point so that it can be determined during shooting, and connect the camera at that position and the specific point on the subject. Using the connecting straight line as the center line, move the camera in a direction perpendicular to the center line an equal distance to the left or right by a distance determined for each shooting position, and move the camera so that the image of the specific point on the subject matches the displayed position. Since the left and right frames of the subject are photographed by adjusting the line of sight, the line of sight always intersects at one specific point on the subject, and the projected 3D image is always located near the screen. You can get an easy-to-view 3D video. Furthermore, since the baseline length for shooting a stereoscopic video can be determined for each shooting position, the stereoscopic effect to be obtained can be arbitrarily selected, and not only a natural three-dimensional effect but also a supernatural three-dimensional effect can be obtained. Furthermore, since only a single movie camera or television camera is used, it is possible to get very close to the subject and shoot a 3D video with a natural 3D effect. Furthermore, it is also possible to capture stereoscopic videos of minute objects.

[Brief explanation of drawings]

Fig. 1 is a diagram for explaining the stereoscopic video shooting method according to the present invention, Fig. 2 is a diagram for explaining the movement of the camera when photographing a subject in close-up with a fixed base line length, and Fig. 3 , FIG. 4 is a diagram for explaining an example of the movement trajectory of the camera, FIG. 5 is a perspective view of an embodiment of equipment for realizing the stereoscopic video shooting method of the present invention, and FIG. 6 is a diagram showing stereoscopic viewing with the naked eye. Figure 7 is a diagram to explain the configuration and operation of a device that uses two conventional cameras to photograph stereoscopic video with binocular parallax, and Figure 8 is a diagram to explain the configuration and operation of a stereoscopic video with binocular parallax using two conventional cameras. FIG. 3 is a diagram for explaining the relationship between the position of a cross point and the position of a stereoscopic image when projected. S...Subject, t...Center line, l...Baseline, P
...Cross point, F... Finder screen,
k...Trajectory, L...Objective lens, 10...Camera, 11...Focus adjustment (focus adjustment) mechanism, 12
...Horizontal angle adjustment mechanism, 13...Vertical angle adjustment mechanism,
14... Horizontal movement mechanism, 15... Height adjustment device,
16... Rail, 17... Mounting table, 18... Camera moving mechanism, 19... Rail, 20... Moving table,
21...Wheel, 22...Subject mounting device, 23...
...Placement table, 24...Rotation adjustment device.

Claims (1)

[Claims] 1. A method for photographing a three-dimensional video of a subject whose external shape does not change at least during photographing, in which a single movie camera or television camera is used, and an image of a specific point on the subject is fixed at a fixed position on the photographing screen. In order to locate the subject, display the image of the subject on the viewfinder or monitor screen of a movie camera or television camera, select a specific point on the subject, display the position of that point so that it can be determined during shooting, and With the straight line connecting the camera at the position and a specific point on the subject as the center line,
Move the camera in one direction perpendicular to the center line by a distance determined for each shooting position, and adjust the line of sight of the camera so that the image of a specific point on the subject matches the displayed position. Take a picture of one frame, move it the same distance in the opposite direction from the center line, adjust the line of sight of the camera so that the image of a specific point on the subject matches the displayed position, and shoot the other frame on the left and right. A method for photographing a three-dimensional video of a subject having a fixed external shape, characterized in that a group of frames for a three-dimensional video are photographed by photographing the above two images at slightly different positions of the camera relative to the subject. 2. The method for photographing a three-dimensional video of a subject having a fixed external shape according to claim 1, wherein the distance determined for each photographing position of the camera is a distance that decreases linearly as the distance from the subject decreases. 3. The fixed external shape according to claim 1, wherein the distance determined for each photographing position of the camera is a distance that decreases along a curve such that the rate of decrease increases in order as the distance from the subject decreases. A method for capturing stereoscopic video of a subject. 4 Adjust the line of sight of the camera so that the image of the specific point on the subject matches the displayed position by measuring the distance between the specific point on the subject and the camera, and using the measured value and the distance the camera moves relative to the center line. 4. The method for photographing a three-dimensional moving image of a subject having a fixed external shape according to claim 1, wherein the direction of the line of sight is calculated based on the calculation. 5. Adjustment of the line of sight of the camera so that the image of the specific point on the subject matches the displayed position is performed while observing the image of the subject on the camera's viewfinder or monitor screen. 3. A method for photographing a stereoscopic video of a subject having a fixed external shape according to any one of Item 3.