JPH11238143A - Image processing method, presentation method of virtual reality, and storage medium for program thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To generate an image capable of presenting virtual reality free of unnaturalness by generating a reduced image of a panoramic image when a current view position is in front of the view position at the time of the panoramic image formation. SOLUTION: When an observer is at a position Z, which is closer to Zn-1 than to a reference position Zn , a panoramic image at the reference position Zn-1 is used. At this time, the current position Z is in front of the position Zn-1 where the panoramic image was photographed. To cut the panoramic image at the position Zn-1 and then enlarge the image, a smaller area is referred to. When the observer is at Z', on the other hand, the observer is close to the reference position Zn . The reference position Zn is positioned ahead of the current position Z', so the observer at the position Z' is far away from a forward scene photographed at the reference position Zn . Therefore, an image cut out of the panoramic image at the reference position Zn is reduced.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of processing an image giving a virtual reality and a method of presenting the virtual image. Such a display method. The invention further relates to a storage medium for a program for presenting virtual reality.

[0002]

2. Description of the Related Art In recent years, attempts have been made to build an urban environment in which people gather and perform social activities in cyberspaces formed by computer networks.
Usually, conventional CG technology is used for describing and displaying this kind of virtual space. However, since there is a limit in CG expression based on a geometric model, ImageBasedRendering (IBR) based on a real image has recently been receiving attention.

[0003] Reproduction of image data obtained by actual photography as it is is merely a re-experience of the photographer. Therefore, based on the image database, an arbitrary landscape is generated and presented in real time as image data using IBR technology. In other words, if each shot video is treated as an independent image and rearranged according to the experience of the user, it will be possible to walk through the unique travel route required by the user in a remote location. You can feel a three-dimensional virtual space there.

Recently, three-dimensional drawing software tools for personal computers have been developed one after another, and three-dimensional walkthrough can be easily performed on personal computers. FIG. 1 to FIG. 4 show display examples of a three-dimensional image in a two-dimensional walkthrough using this software tool. In FIG. 1, the observer can move in the Z-axis and X-axis directions. It is also assumed that the image database 100 previously stores a real image or a CG image at a plurality of viewpoint positions in the Z-axis direction, and stores a panoramic image having a substantially 360-degree spread at each viewpoint position.

For the sake of simplicity, the scene viewed by the user is, as shown in FIG. 2, a point A in the Z-axis direction, which is the plus direction of the X-axis (θ = 90 ° direction).
The triangle that exists in the “+ X direction”) and the point A
It is assumed that the sphere exists in the minus direction of the X axis (θ = −90 degrees direction, hereinafter referred to as “−X direction”). FIG. 3 shows a panoramic image 200 that has been previously photographed of the scene of FIG. 2 at a certain point on the Z axis, converted into a panoramic image, and stored as an image database. This panoramic image 200 is obtained by transforming a triangular image 210 in a direction of θ = 90 degrees (= + X direction) into a direction of θ = 270 degrees (= + X direction).
(-X direction).

The conventional software tool described above cuts out a scene having a predetermined angle of view in the direction of θ = 90 degrees from the panoramic image 200 when the observer looks in the + X direction at the point A, for example, by using the first As shown in the figure, the scene 201 is presented to the observer. As shown in FIG.
1 includes a triangular shape 211. Here, when the observer moves from point A by a distance Δx (assuming Δx> 0) to position B in the + X direction as shown in FIG. 4, this software tool changes the angle of view to The panorama image is cut out by changing it, that is, the image is cut out in a range narrower than the angle of view cut out in FIG. 1, and the cut out image is reduced to the same size as the image 201 in FIG. It is enlarged and presented as a scene 202 as shown in FIG. The scene 202 includes a triangular shape 212 (enlarged from 211 in FIG. 1).

[0007] The observer observes this enlarged triangular shape 212.
Is presented, the user feels a pseudo proximity (movement).

[0008]

The conventional three-dimensional software tool has the following disadvantages. If the observer rotates the line of sight by 180 degrees at position B, position B is supposed to be reduced by ΔX from the object. The sphere (circular figure) is enlarged and displayed, giving an uncomfortable feeling to the observer.

That is, as shown in FIG. 5, when the observer looks back at point B, that is, when the observer turns in the −X direction, there is a spherical object in that direction. A scene in the direction of θ = −90 degrees is cut out from 200 and is presented to the observer. In this conventional tool, the observer uses the distance Δx in the + direction from the Z axis.
Is stored, so that the above-described image processing (that is, enlargement) performed on the scene 202 due to the movement of Δx in the + direction is changed to a scene in the θ = −90 degree direction. As shown in FIG. 5, a circular figure which should be reduced originally is enlarged and displayed like 221 as shown in FIG. Originally,
Since the observer has moved from the Z axis in the + X direction by the distance x, the observer should have moved away from the sphere by the distance x. However, the observer mistakenly treats the distance Δx as approaching. .

The drawback of this three-dimensional software is that, as shown in FIG. 6, the observer must change his line of sight in order around point B and finally turn to the −X direction. This is because the tool approximates the line-of-sight direction in the −X direction to the line-of-sight direction at the point B ′, as shown in FIG. This is because in a simple system such as a personal computer, to realize a real-time walkthrough,
This is because the above approximation is necessary.

[0011]

SUMMARY OF THE INVENTION The present invention has been proposed to improve the drawbacks of a virtual reality presentation system using a panoramic image as represented by the above-mentioned conventional three-dimensional software tool. The purpose of the present invention is to propose an image processing method for generating an image giving a virtual reality without a sense of incongruity, and a method for presenting a virtual reality using the image.

In order to achieve the above object, the image processing method according to the present invention proposes a presentation method that gives a virtual reality without discomfort when the observer looks back, for example. For this purpose, an image for generating a virtual image that gives an observer a sense of movement accompanying a rotational movement based on a panoramic image created so as to have a plurality of line-of-sight directions from one viewpoint position according to claim 1 The processing method is such that, when the observer turns substantially from the first line of sight to a second line of sight substantially 180 degrees opposite to the first line of sight, the current viewpoint position is changed to the panoramic position. A generating step of generating a reduced image from the panoramic image when the image is located ahead of the viewpoint position at the time of creating the image in the first line-of-sight direction, and an enlarged image when the image is located backward. It is characterized by doing.

When the observer substantially looks back, if the viewpoint position is ahead of the viewpoint position when the panoramic image is created in the first line-of-sight direction, the scene in the turned direction is It is possible to eliminate a sense of incongruity by reducing the size of the scene because it is far away, and by enlarging it when it is near the scene when it is behind.

The present invention is directed not only to the movement of the observer turning around (rotating movement) but also to the elimination of discomfort when performing the parallel movement movement. In order to achieve this object, the image processing method according to claim 2, wherein the observer has the first
When rotating while translating in a direction substantially orthogonal to the line of sight, detecting the amount of parallel movement and shifting the panoramic image based on the detected amount of parallel movement to generate an image. It also has In other words, the parallel movement of the observer is such that the virtual reality does not feel strange due to the shift processing of the panoramic image.

According to a preferred aspect of the present invention, in the shift step, the step of obtaining a cutout position based on the detected amount of parallel movement and the step of cutting out an image around the cutout position By having
Achieve image shifting. Another object of the present invention is to eliminate a sense of incongruity when the observer moves in the direction of arrangement of the panoramic images and the line of sight of the observer is substantially along the direction of arrangement.

[0016] To this end, an image processing method according to a fourth aspect is characterized in that the panoramic image is composed of one panoramic image in a panoramic image sequence from a plurality of discrete viewpoint positions arranged in a predetermined sequence direction. There is further provided a moving position detecting step of obtaining a moving position of the observer in the series direction, and a panoramic image specifying step of specifying a panoramic image at a viewpoint position closest to the moving position. It is characterized in that a panoramic image is used as an image on which enlargement or reduction is performed in the generation step.

According to a preferred aspect of the present invention, the enlargement or reduction magnification is based on a distance vector from the viewpoint position along the first line of sight. According to claim 6, which is a preferred aspect of the present invention, when the current line-of-sight direction is oblique to the direction of the panoramic image sequence, enlargement / reduction processing and shift processing are performed.

In order to present virtual reality, it is preferable that the sizes of the presented images do not differ. Therefore, the image generated by the image processing method is corrected to a predetermined size according to claim 7 and presented to the observer. An object of the present invention is to propose a presentation method to which a technique based on a panoramic image is applied when a virtual reality image without a sense of incongruity can be presented.

To this end, a virtual reality system for presenting a virtual image that gives a sense of movement to an observer by using a series of panoramic image sequence images from a plurality of discrete viewpoint positions according to claim 8. Is selected according to the current viewpoint position, either the first method based on the panoramic image according to the current viewpoint position or the second method that can be freely moved two-dimensionally. A virtual reality image at the current viewpoint position is generated by the selected method, and the generated image is presented to the observer.

According to a preferred embodiment of the present invention, the second method generates a real environment image based on a ray space theory. Both the image processing method and the virtual reality presentation method having the above configuration can be computer-programmed,
Therefore, it can be stored in a storage medium.

[0021]

BRIEF DESCRIPTION OF THE DRAWINGS FIG.
A virtual reality presentation system according to a preferred embodiment of the present invention will be described. FIG. 8 shows the hardware configuration of this system. This system is built on an ordinary workstation based on WINDOWS NT or UNIX. For example, a CPU 13 using a microcomputer such as PENTIUM II reads application programs stored in a hard disk 14 onto a RAM 11 and By executing, a virtual reality image is constructed and CR
It is displayed and realized on T12.

An image for presenting virtual reality is generated by cutting out a panoramic image stored in the hard disk 15 in advance. As shown in FIG. 9, the panoramic image sequence is composed of a plurality of generated panoramic images previously captured at a plurality of discretely set reference positions. In the example of FIG. 9, one panoramic image is
At one reference position, the image is obtained from eight real shot images taken in discrete directions, for example, in eight directions. That is, one panoramic image is composed of its reference position, azimuth, and image data.

The panoramic image of this embodiment has a columnar space as shown in FIG. As shown in the figure, an arbitrary scene is generated by cutting out an image in a direction corresponding to the viewing direction. In the present embodiment, in order to present a distant scene, a wide range of image data is cut out from the panoramic image, and a reduced version of the image data is presented, so that the sense of separation is presented in a pseudo manner. However, if a wider range is cut out in the vertical direction, the vertical deviation is more emphasized. Therefore, the panoramic image database of the present embodiment has an elevation angle Φ as shown in FIG.
Image data 200 of a panoramic image at ₀ (= 0 degree)
In addition, image data (300, 400) of a panoramic image at an elevation angle between Φ ₁ (> 0) and Φ ₂ (<0) is also stored.

When the observer views the scene by moving backward, that is, when it is necessary to cut out an image by increasing the reduction ratio, that is, when it is necessary to cut out a wide range of image data, the cutout range becomes panoramic. Image 30
In the case of 0 and / or panoramic image 400, image data is cut out from panoramic image 300 and / or panoramic image 400.

The actual viewpoint position and line-of-sight direction are
As shown in the figure, the reference position (Z ₁ , Z ₂ , Z ₃ ...) And the photographing direction (θ ₁ , θ ₂ , θ ₃ ...) Are different. In FIG. 12,
The current viewpoint position has (x, z, θ), that is, the viewpoint position (x, z) is Δx, Δ from the closest reference position Z _0.
It is shifted by z. In addition, the line-of-sight direction θ shooting direction θ ₁ ~
θ ₈ is different.

FIG. 13 shows a control procedure of the application program of the embodiment. A procedure for generating a virtual reality image will be described with reference to FIG. 13 and other drawings. In step S2, entering a current viewpoint position _{(x n, z n, θ} n). In step S4, find the nearest reference position Z _n to the viewpoint position coordinate z _n. The Z _{n is, Z n: MIN k {(} x n -X k) 2 + (z n -Z k) 2} is determined from.

The call In step S6, the panoramic image PI associated with the reference position Z _n a (n) from the disk 15. In step S8, the "shift amount" is determined. In the present embodiment, the image is shifted in order to generate a virtual reality image from a line of sight shifted from the position and direction in which the panoramic image was captured. In FIG. 14, a solid-line rectangle indicates an area of the cut-out panoramic image, and a wavy-line rectangle indicates a CRT.
13 shows a display area displayed. If the wavy rectangle is displayed on the CRT 13 as it is, it means that an image from the shooting position of the panoramic image is presented as a virtual reality image. On the other hand, the current viewpoint position is ahead reference position Z _n (shifted to the right, i.e., Delta] z> 0)
In this case, as shown in FIG.
1 is moved to the right as indicated by 502. This image 502
There When displayed, the viewer is presented as one of such image viewed from a position shifted to the right by Δz from the reference position Z _n, i.e., pseudo moving feeling is given. Also,
Current viewpoint position (shifted to the left, i.e., Delta] z <0) which is delayed from the reference position Z _n if, as shown in FIG. 15, the rectangular region 501 of the wavy line as 503 Move to the left. When this image 503 is displayed, the viewer is presented as one of such image viewed from a position shifted to the left by Δz from the reference position Z _n.

Returning to the control procedure of FIG.
The “shift amount” of 8 is the movement amount Δz of the viewpoint position in the Z-axis direction.
The shift amount is determined according to. Since this shift amount is a function of the movement amount Δz, it is represented by f (Δz). Step S1
At 0, the “enlargement / reduction rate g” is calculated. As shown in FIG. 16, when modeling is performed by assuming that the line-of-sight direction is θ = 90 ° and Δz = 0, if the viewpoint position shifts by Δx from the Z axis in the X direction, if Δx> 0, You will see an image of a scene at θ = 90 degrees, but since it is approaching this scene by Δx, the virtual reality image must be enlarged. On the other hand, when Δx <0, Δx
Since it is only far away, the virtual reality image must be reduced.

When the viewing direction is -90 degrees, on the contrary, a panoramic image of θ = -90 degrees is viewed,
If Δx> 0, the virtual reality image is reduced by Δx from the landscape, and if Δx <0, the virtual reality image is enlarged by Δx. It must be. Accordingly, the enlargement or reduction magnification g depends on the absolute value of the separation distance Δx from the Z-axis, and whether the enlargement or reduction is performed depends on whether Δx approaches or moves away from the scenery as shown in FIG. ,
Δx). In step S10, the enlargement / reduction magnification is determined in this way.

The shift amount f and the enlargement ratio g calculated in steps S8 and S10 are calculated by modeling the visual line direction θ as 90 degrees, but the visual line direction is not actually 90 degrees. Therefore, in step S12, the shift amount f and the enlargement ratio g are corrected. That is, the shift amount f
Is corrected by f × (π / 2−θ) / (π / 2), and the magnification ratio g is corrected by g × (θ) / (π / 2).

Subsequently, in step S14, step S
From the panoramic image PI (n) extracted in step 6
An image is cut out based on the shift amount f and the enlargement / reduction magnification g calculated in steps 8 and 10. Here, the position of the image to be clipped is determined by the shift amount f. The size of the image to be cut out is 1 / g of the display image size.

When the range to be cut out is large, as shown in FIG. 11, in order to reduce distortion or positional deviation, in addition to panoramic image 200, panoramic image 300 and / or panoramic image 400 are used. Should also be cut out. In step S14, the clipped image is enlarged by a factor of g. That is, the image cut out at the size of 1 / g in step S12 is enlarged by g times in step S16, so that the enlarged image matches the size of the display screen. In the eyes, a scene approaching or moving away is displayed in a pseudo manner.

FIG. 17 shows that the rectangular area 504 of the wavy line
In contrast, in FIG. 18, the dashed rectangular area 505 is enlarged with respect to 501 to reduce the size of the display graphic. Each is shown to be reduced. Incidentally, the shift amount f and the enlargement / reduction ratio g in step S12.
Can be corrected with higher accuracy. That is, when the correction amount is calculated with high accuracy, as shown in FIG.

[0034]

(Equation 1)

On the other hand, for the enlargement ratio, θ
Where the amount of proximity to the direction is:

[0036]

(Equation 2)

Calculated based on FIG. 20 to FIG.
FIG. 2 schematically shows the range and direction of a cut-out image when an image giving virtual reality is generated by the control procedure of FIG. In the drawing, the hatched sector indicates a reference range, and the smaller the angle of the sector, the larger the enlargement ratio, and the larger the angle of the sector, the larger the reduction ratio.

That is, FIG. 20 shows the enlargement / reduction in the case where the line of sight moves in an upward direction or the downward direction in the drawing from the Z axis when the line of sight is indicated by each sector. FIG. 21 shows the case where the moving direction is oblique, as shown in FIG.
FIG. 2 shows a case where the observer rotates on the spot. A conventional virtual reality presentation system, for example, when an observer turns around like the three-dimensional image software tool described above in the section of the prior art, is shown in FIG. 1 and FIG. Although an unnatural image is presented as described above, in the present embodiment, as is clear from the control procedure of FIG. 13 and FIG. 22, a virtual reality image that makes a natural and smooth change is presented.

The enlargement / reduction of the control procedure shown in FIG. 13 is effective even in the case where the panorama image sequence proceeds in the arrangement direction. FIG. 23 shows a case where the viewing direction of the observer's traveling direction (that is, the direction of the panoramic image sequence) is the same.
FIG. 24 shows the case of the reverse direction. FIG. 23, second
In FIG. 4, Z indicates the current position of the observer at a certain time, and Z 'indicates the current position of the observer at another time.

In FIG. 23, the observer is at position Z.
If so, this Z is the reference position Z _nZ than_n-1Close to
The reference position Z_n-1Use panoramic images of
Will be. At this time, the current position Z takes a panoramic image
Position Z_n-1Exists ahead of it. That is, the pano
The observer has advanced to the scene ahead of the Lama image
You will be in Therefore, the observer is
Because it is closer, you should see the enlarged image
is there. Therefore, the observer is located at Z in FIG.
In the case, the position Z_n-1Panorama image color image clipping
To make the image larger
It refers to the area.

On the other hand, in FIG. 23, when the observer is located Z 'is, the viewer is close to the reference position Z _n. 'Since the position advanced than the position Z' reference position Z _n is the current position Z observer will be at the more distant position from the front scene taken at the reference position Z _n. Therefore, when the observer is in the Z 'is to reduce the clipped image from the panoramic image at the reference position Z _n.

FIG. 24 shows a case where the observer is looking in the opposite direction to the traveling direction. Enlargement / reduction of the cut-out image in FIG. 24 is opposite to that in FIG. Presentation of virtual reality based on a panoramic image according to the present invention is not always generated based on a panoramic image. For example, as shown in FIG. 25, areas (areas A and A ′) where walk-throughs proceed in substantially one direction (for example, one-dimensional direction).
When there is an area B that performs a two-dimensional movement, for the areas A and A ′, a real image is presented based on the above-described panoramic image, or on the other hand, for the area B, In this case, a technique that can move freely in two dimensions is applied. As a technique that can freely move two-dimensionally, for example, there is a technique that applies a ray space theory.

As shown in FIG. 26, for example, the areas A and A 'include a road area mainly for walking. On the other hand, in the area B, there is a park (or a building) that it is usual to walk around. Therefore, as shown in FIG. 27, a flag is added to each region in advance as to whether the method based on the panoramic image is suitable or the region based on the ray space theory method. Which area is located is determined by the value of the flag, and the method is switched according to the value of the flag.

The present invention can be further modified. In the above embodiment, the panorama image uses a photographed image, but C
It may be generated from a G image. Further, even if the Z-axis direction, that is, the moving direction of the observer is not a straight line or a free curve, the present invention is not affected as long as a panoramic image sequence is provided along the free curve. .

[0045]

As described above, according to the present invention, it is possible to generate an image that gives the observer a virtual reality without discomfort based on the panoramic image.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a drawback of a conventional virtual reality presentation system, and is a diagram illustrating generation of a virtual reality image when an observer is on a Z axis.

FIG. 2 is a diagram illustrating an example of a panoramic image for explaining a defect of a conventional virtual reality presentation system.
A triangular pyramid and a sphere are arranged across the axis.

FIG. 3 is a diagram showing an example of a panoramic image of a conventional virtual reality presentation system.

FIG. 4 is a diagram for explaining a drawback of the conventional virtual reality presentation system, and is a diagram showing generation of a virtual reality image when an observer moves away from the Z axis.

FIG. 5 is a diagram for explaining the drawbacks of the conventional virtual reality presentation system, showing generation of a virtual reality image when an observer looks back.

FIG. 6 is a diagram for explaining the drawbacks of the conventional virtual reality presentation system, showing generation of a virtual reality image when an observer looks back.

FIG. 7 is a diagram for explaining the drawbacks of the conventional virtual reality presentation system, and is a diagram illustrating generation of a virtual reality image when an observer looks back.

FIG. 8 is an exemplary view for explaining the hardware configuration of the system according to the embodiment;

9 is a diagram showing a configuration of a panoramic image sequence used in the system of FIG.

FIG. 10 is an exemplary view for explaining a method of cutting out a display image from a panoramic image in the system according to the embodiment.

FIG. 11 is an exemplary view for explaining that a panoramic image used in the system according to the embodiment requires a plurality of panoramic images according to a change in the vertical direction of the line of sight.

FIG. 12 is a diagram illustrating a state in which the current position of the observer is shifted from the reference position Z ₀ by (Δx, Δz, θ) in the embodiment.

FIG. 13 is a flowchart illustrating a control procedure according to the embodiment.

FIG. 14 shows a case where the observer translates rightward.
FIG. 7 is a diagram for explaining a state in which right shift processing is performed on an image cut out from a panoramic image.

FIG. 15 shows a case where the observer translates to the left.
FIG. 9 is a diagram illustrating a state in which a left shift process is performed on an image cut out from a panoramic image.

FIG. 16 is a view for explaining the reason why a panoramic image is enlarged or reduced when an observer moves in a direction perpendicular to the Z axis.

FIG. 17 is a view for explaining the reason for enlarging the panoramic image when the observer moves close to the panoramic image in a direction perpendicular to the Z axis.

FIG. 18 is a view for explaining the reason for reducing the panoramic image when the observer moves away from the panoramic image in a direction perpendicular to the Z axis.

FIG. 19 is a view for explaining the principle of calculating a shift amount and an enlargement magnification with high accuracy.

FIG. 20 is a diagram illustrating an operation example based on the control procedure of the embodiment, and is a diagram illustrating a reference direction and a reference angle of a panoramic image when an observer moves in a direction perpendicular to the Z axis.

FIG. 21 is a diagram illustrating an operation example based on the control procedure of the embodiment, and is a diagram illustrating a reference direction and a reference angle of a panoramic image when an observer moves in an oblique direction with respect to the Z axis.

FIG. 22 is a diagram illustrating an operation example based on the control procedure of the embodiment, and is a diagram illustrating a reference direction and a reference angle of a panoramic image when an observer rotates.

FIG. 23 is a diagram illustrating that the system according to the embodiment is also effective when moving in a line of sight parallel to the traveling direction.

FIG. 24 is a diagram illustrating that the system according to the embodiment is also effective when moving in a line of sight parallel to the traveling direction.

FIG. 25 is a diagram illustrating a method of presenting virtual reality according to another embodiment, and is a diagram illustrating a method of presenting virtual reality by switching between a panoramic image technique and another technique.

FIG. 26 is a view for explaining an example to which the method of FIG. 25 is applied.

FIG. 27 illustrates an example to which the method of FIG. 25 is applied.

Claims (16)

[Claims] 1. An image processing method for generating, based on a panoramic image created so as to have a plurality of line-of-sight directions from one viewpoint position, a virtual image that gives an observer a sense of movement accompanying a rotational movement. When the observer looks back substantially from the first line of sight to a second line of sight substantially 180 degrees opposite to the first line of sight, the current viewpoint position is A generating step of generating an image reduced from the panoramic image when the image is located ahead of the viewpoint position at the time of creation in the first line-of-sight direction, and generating an enlarged image when the image is located backward. An image processing method characterized by the following. 2. The method according to claim 1, wherein when the observer rotates while translating in a direction substantially orthogonal to the first line-of-sight direction, a step of detecting a translation amount, and the panorama based on the detected translation amount. The image processing method according to claim 1, further comprising a shift step of generating an image obtained by shifting the image. 3. The shifting step comprises: a step of obtaining a cutout position based on the detected parallel movement amount; and a step of cutting out an image centered on the cutout position. 3. The image processing method according to claim 1, wherein: 4. The panoramic image is one panoramic image in a panoramic image sequence from a plurality of discrete viewpoint positions arranged in a predetermined sequence direction, and is a movement position of an observer along the sequence direction. And a panoramic image specifying step of specifying a panoramic image at a viewpoint position closest to the moving position. The panoramic image is used as a basis for enlargement or reduction in the generation step. The image processing method according to claim 1, wherein the image processing method is used as an image. 5. The image processing apparatus according to claim 1, wherein the enlargement or reduction magnification is based on a distance vector from the viewpoint position along the first line-of-sight direction. Method. 6. The image processing method according to claim 4, wherein when the current line-of-sight direction is oblique to the series direction, the enlargement / reduction processing and the shift processing are performed in combination. 7. A presentation of virtual reality, wherein an image reduced or enlarged by the method according to any one of claims 1 to 6 is corrected to a predetermined size and presented to an observer. Method. 8. A virtual reality presentation method for presenting a virtual image that gives a sense of movement to an observer using images of a series of panoramic image sequences from a plurality of discrete viewpoint positions, comprising: Either the first method based on the panoramic image or the second method that can be freely moved two-dimensionally according to the position is selected according to the current viewpoint position, and the current viewpoint is selected according to the selected method. A virtual reality presentation method characterized by generating a virtual reality image at a position and presenting the generated image to an observer. 9. The virtual reality presentation method according to claim 7, wherein the second method generates a real environment image based on a ray space theory. 10. An image processing program for generating, based on a panoramic image created so as to have a plurality of line-of-sight directions from one viewpoint position, a virtual image that gives the observer a sense of movement associated with a rotational movement. When the observer looks back substantially from the first line of sight to a second line of sight substantially 180 degrees opposite to the first line of sight, the current viewpoint position is Generating an image reduced from the panoramic image if the image is located ahead of the viewpoint position at the time of creating the panoramic image in the first line of sight direction, and generating an enlarged image if the image is located behind the image. A storage medium for storing a code. 11. The detection program code for detecting a translation amount when an observer rotates while translating in a direction substantially perpendicular to the first line-of-sight direction, the detection program code comprising: The storage medium according to claim 10, further comprising: a shift program code for generating an image obtained by shifting the panoramic image based on a movement amount. 12. The shift program code further stores: a program code for obtaining a cutout position based on the detected parallel movement amount; and a program code for cutting out an image around the cutout position. Claim 10 or 11
A storage medium according to claim 1. 13. The panoramic image is one panoramic image in a panoramic image sequence from a plurality of discrete viewpoint positions arranged in a predetermined sequence direction, and is a moving position of the observer in the sequence direction. And a panoramic image specifying program code for specifying a panoramic image at a viewpoint position closest to the moving position.
A storage medium according to any one of claims 1 to 3. 14. The method according to claim 1, wherein the enlargement or reduction magnification is set to the first magnification.
The storage medium according to any one of claims 10 to 13, further storing a program code that is calculated based on a distance vector from the viewpoint position along the line of sight direction. 15. The method according to claim 13, further comprising storing a program code for performing both enlargement / reduction processing and shift processing when the current line-of-sight direction is oblique to the sequence direction. The storage medium according to the above. 16. The storage medium according to claim 10, further storing a program code for presenting the generated image.