JP3544752B2 - Image information generation method, image display method, and image display system - Google Patents

Description

[0001]
[Industrial applications]
The present invention relates to a method of generating stereoscopically viewable image information, an image display method using the image information, and an image display system.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, methods for generating image information that can be viewed stereoscopically include methods such as random dot stereogram (RDS) and color field stereogram. In the random dot stereogram method, a plurality of adjacent modified images are generated from a basic (base) pattern image. At this time, the basic pattern image is formed by randomly arranging images of a predetermined color and a predetermined luminance in pixel (pixel, dot) units. On the other hand, a plurality of adjacent modified images are generated from the basic pattern image also in the color field stereogram method, and the basic pattern image is formed by drawing a pattern or the like of a predetermined shape in pixel units.
[0003]
In any of the above methods, the sense of depth of the generated virtual image is determined by the distance between the left eye point and the right eye point. For example, in stereoscopic viewing by the intersection method shown in FIG. 12A, the depth value of the virtual image decreases as the distance between the gazing points increases (assuming that the depth value increases toward the back side toward the screen). That is, in FIG. 12A, C1 is a virtual image generated by the pixel A at the right eye point and the pixel B1 at the left eye point, and C2 is generated by the pixel A and the pixel B2 at the left eye point. It is a virtual image to be performed. Since the distance between A and B2 is larger than that between A and B1, the depth value of the virtual image C2 is smaller than C1. On the other hand, in the stereoscopic view by the parallel method shown in FIG. That is, in FIG. 12B, F1 is a virtual image generated by the pixel D at the left eye point and the pixel E1 at the right eye point, and F2 is generated by the pixel D and the pixel E2 at the right eye point. It is a virtual image to be performed. Since the distance between D and E2 is greater than that between D and E1, the depth value of virtual image F2 is greater than F1. As described above, in the random dot stereogram method or the like, a sense of depth is created by controlling the magnitude of the distance between gazing points, and an image that can be stereoscopically viewed is generated.
[0004]
[Problems to be solved by the invention]
However, in the conventional method, the size of the distance between gazing points can be controlled only in pixel units, and the size is limited to an integral multiple of the pixel size. For this reason, a possible depth value is a discrete value, and there has been a problem that the obtained virtual image has a hierarchical (step-like) shape in the depth direction (see FIG. 6A). This problem is not so great when an image is created on the premise that the image is printed by a high-resolution printer, for example. However, when the obtained image is displayed on a display device (television, display, or the like) having a finite number of pixels, the step-like shape is conspicuous and gives an unnatural feeling to a viewer. Therefore, a technical problem is how to solve this problem.
[0005]
The present invention has been made in view of the above problems, and has as its object to provide an image information generation method, an image display method, and an image information generation method that can improve the quality of an image generated by a random dot stereogram method or the like. An image display system is provided.
[0006]
Means and Action for Solving the Problems
In order to solve the above problem, the invention according to claim 1 is based on one of left-eye and right-eye gazing point image information and another gazing point distance determined by depth information. An image information generating method for generating image information in pixel units that can be stereoscopically viewed by determining the point of interest image information,
Obtaining the one point-of-interest image information in sub-pixel units, and the other point-of-interest image information in sub-pixel units based on the inter-point-of-regard distance set in sub-pixel units;
Obtaining image information in pixel units from the other point-of-interest image information in sub-pixel units.
[0007]
In addition, the invention of claim 11 is based on one of the point of interest image information for the left eye and the right eye, and the distance between the points of interest determined by the depth information, the other point of interest image information different from the one An image display system that generates image information in pixel units that can be obtained stereoscopically and displays the image information on a display device having a finite number of pixels,
Means for obtaining the other point-of-interest image information in sub-pixel units, based on the one point-of-interest image information in sub-pixel units and the distance between points of interest set in sub-pixel units,
Means for obtaining image information in pixel units from the other point-of-interest image information in sub-pixel units.
[0008]
According to the first or eleventh aspect of the present invention, the other gazing point image information in the sub-pixel unit is obtained based on the distance between gazing points set in the sub-pixel unit, and is converted into image information in the pixel unit. Therefore, the distance between the gazing points can be controlled in units of sub-pixels, and it is possible to obtain a stereoscopic image that generates a more accurate and smooth virtual image having an inclined shape. The present invention can be applied not only to the random dot stereogram / color field stereogram method but also to various stereoscopic image information generating methods.
[0009]
Further, the invention according to claim 2 forms a plurality of adjacent modified image information based on the basic pattern image information and the distance between gazing points determined by the depth information, and converts the image information in pixel units that can be stereoscopically viewed. A method for generating image information, comprising:
Converting the basic pattern image information in pixel units into image information in subpixel units;
Based on at least one of the basic pattern image information in sub-pixel units, the (N-1) th modified image information (N is an integer) in sub-pixel units, and the distance between gazing points set in sub-pixel units, Generating N modified image information;
Obtaining image information in pixel units from the obtained modified image information in subpixel units.
[0010]
According to the second aspect of the present invention, the N-th modified image in sub-pixel units is successively obtained based on the basic pattern image information, the (N-1) -th modified image, the distance between gazing points in sub-pixel units, and the like. , Are converted into image information in pixel units. This makes it possible to generate a stereoscopic image based on the basic pattern image, and to obtain a stereoscopic image that generates a more accurate and smooth virtual image having a tilted shape. As a method for obtaining image information in pixel units from image information in subpixel units, various methods such as performing γ correction at the time of conversion in addition to a method for obtaining an average value can be adopted.
[0011]
According to a third aspect of the present invention, in the method of any one of the first and second aspects, in the step of obtaining the image information of the pixel unit, an average value of image information of sub-pixels included in the pixel is set as the image information of the pixel unit. It is characterized by the following.
[0012]
According to the third aspect of the present invention, the image information of the pixel can be obtained as an average value of the image information of all the sub-pixels included in the pixel. For example, image display on a display device having a finite number of pixels can be performed. It becomes.
[0013]
According to a fourth aspect of the present invention, in the second aspect, the basic pattern image information is composed of sub-pixel unit image information, and in the step of obtaining the pixel unit image information, It is characterized in that image information in pixel units is obtained from basic pattern image information.
[0014]
According to the fourth aspect of the present invention, the basic pattern image information is composed of image information in units of sub-pixels from the beginning, and there is no need to perform processing for converting the image information into image information in units of sub-pixels.
[0015]
According to a fifth aspect of the present invention, in any one of the second to fourth aspects, the basic pattern image information is formed by arranging images randomly in pixel units or subpixel units.
[0016]
According to the fifth aspect of the present invention, it is possible to obtain a random dot stereogram image that generates a virtual image having an accurate and smooth inclined shape.
[0017]
According to a sixth aspect of the present invention, in any one of the second to fourth aspects, the basic pattern image information is formed by drawing a pattern of a predetermined shape in pixel units or subpixel units.
[0018]
According to the sixth aspect of the present invention, it is possible to obtain a color field stereogram image that generates an accurate and smooth virtual image having an inclined shape.
[0019]
According to a seventh aspect of the present invention, there is provided the image processing apparatus according to any one of the fourth to sixth aspects, wherein at least one of the vertical and horizontal dimensions of the region where the image information is the same is a non-integer multiple of the pixel size. Is formed.
[0020]
According to the invention of claim 7, it is possible to display an image with uniform brightness even in a display device or the like in which γ correction has not been correctly performed.
[0021]
According to an eighth aspect of the present invention, in any one of the fourth to sixth aspects, the basic pattern image information is formed such that a region where the image information is the same always crosses a pixel boundary.
[0022]
According to the invention of claim 8, it is possible to display an image with uniform brightness even in a display device or the like in which γ correction has not been correctly performed.
[0023]
According to a ninth aspect of the present invention, the image information generated by the image information generating method according to any one of the first to eighth aspects is displayed on a display device having a finite number of pixels.
[0024]
According to the ninth aspect of the present invention, it is possible to display a stereoscopic image that generates an accurate and smooth virtual image having an inclined shape on a display device having a finite number of pixels.
[0025]
According to a tenth aspect, in the ninth aspect, the image information of the moving image is displayed on the display device.
[0026]
According to the tenth aspect, it is possible to obtain a moving image that can be viewed stereoscopically, and it is possible to effectively prevent a situation in which a virtual image is stepped depending on the direction of a display object or the like.
[0027]
【Example】
Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.
[0028]
(First embodiment)
FIG. 1 shows an example of a random dot stereogram (hereinafter referred to as RDS) image generated according to the present embodiment. The RDS image is composed of a basic pattern image and a plurality of (first to fourth) adjacent modified images as shown in FIG. In the basic pattern image, images are randomly arranged in pixel units (pixels of different colors or luminance are arranged at random). Therefore, even in a modified image generated based on this basic pattern image, a random image is generated in pixel units. The image is arranged in the. In order to generate such an RDS image, for example, first, a three-dimensional shape to be expressed is prepared as shown in FIG. Then, depth information is obtained from the three-dimensional shape. For example, in FIG. 2, the depths of the sphere 10 and the cube 12 are arranged before the wall 14, and thus the obtained depth information also indicates that. That is, in the present embodiment, the depth value of the object on the near side with respect to the screen is reduced, for example, and the depth value of the object on the rear side is increased, for example (the reverse is also possible). However, when obtaining depth information, it is not always necessary to prepare a three-dimensional shape as shown in FIG. 2, and depth information may be directly generated using a predetermined program or the like.
[0029]
In the RDS, as shown in FIG. 1, modified images are successively generated from depth information and a basic pattern image prepared in advance. FIG. 3 shows a conventional method for obtaining a modified image from a basic pattern image. Now, assume that the horizontal width of the basic pattern image is 80 pixels. When obtaining image information of a modified image, first, a depth value is referred to, and a distance between gazing points is calculated from this value. In the RDS, the depth value determines the size of the distance between the gazing points. For example, in the parallel method (see FIG. 12B), the distance between the gazing points on the deeper side (the one with a larger depth value) is The distance between the fixation points becomes smaller at a position closer to the front side (a value having a smaller depth value) (the relationship is reversed in the intersection method). As shown in FIG. 3, when the depth value of the first pixel of the first modified image is 0, and the distance L1 between gazing points is 80 pixels with respect to this, as shown in FIG. The image information C1 at the position is read. Thus, the image information (color, luminance, etc.) of the first pixel of the first modified image becomes C1. Next, referring to the depth value of the second pixel, it is −1, and the distance between gazing points is 79 pixels. As a result, the image information C3 on the left side by 79 pixels is read out, whereby the image information of the second pixel becomes C3. Since the distance L3 between the fixation points of the third pixel is 78 pixels, the image information of the third pixel is C5. Similarly, the image information of the fourth, fifth, and sixth pixels is C4, C5, and C7, respectively. Thus, the image information of the pixel of the first modified image is obtained, and the second and third modified images are obtained. Etc. are also required sequentially.
[0030]
As is clear from the above, in the conventional method, the magnitude of the distance between the gazing points could be controlled only in pixel units. If the size of the distance between fixation points can be controlled only in pixel units, the possible value of the distance to the virtual image is limited accordingly. For example, in FIG. 3, a case is considered where the depth value of the first pixel is −0.5 and the distance between gazing points corresponding to the depth value is 79.5 images. In this case, the value of the distance L1 between fixation points that can be actually taken is either 79 pixels or 80 pixels. Therefore, either the image information C1 of the first pixel or the image information C2 of the second pixel of the basic pattern image is selected as the image information of the first pixel of the first modified image. This means that after all, even if the depth value is changed in the range of -1 to 0, the distance to the virtual image is only a value corresponding to one of the depth values 0 and -1. As described above, since the possible values of the distance to the virtual image are discrete, a problem occurs that the obtained virtual image looks like a step (see FIG. 6A).
[0031]
In this embodiment, in order to solve the above problem, first, as shown in FIG. 4, an image (basic pattern image) expressed in pixel units is converted into an image in subpixel units. For example, in FIG. 4, one square pixel is divided into 16 square sub-pixels. However, how to divide one pixel is completely arbitrary, and the number of divisions may be increased or decreased as compared with the case of FIG. 4, or may be divided into figures other than squares such as rectangles and rhombuses. Absent.
[0032]
Next, the generation of the modified image from the basic pattern image is performed not based on the pixel unit as in the conventional method, but on the basis of the inter-gazing point distance set in the sub-pixel unit as shown in FIG. For example, in FIG. 5, when the basic length is 80 pixels, the size of the distances L1, L2, and L3 between the gazing points is 79.75 pixels. As described above, according to the present embodiment, the distance between the gazing points can be controlled in subpixel units, for example, in 0.25 pixel units. Then, the first modified image is generated based on the basic pattern image and the distance between the gazing points in sub-pixel units (for example, the image information of the sub-pixel S2 is located at a position separated from S2 by the distance between gazing points L1). It becomes the same as the image information of a certain sub-pixel S1).
[0033]
Similarly, the N-th (N is an integer) modified image is generated based on the (N-1) -th modified image (or the basic pattern image) and the distance between gazing points in sub-pixel units. As described above, according to the present embodiment, the distance between fixation points is in units of sub-pixels, so that a stereoscopic image that generates a more accurate and smooth virtual image with an inclined shape can be obtained as compared with the conventional example. That is, in the present embodiment, the distance to the virtual image, which could only be changed by one unit in the conventional example, can be changed in, for example, a unit of 0.25. The image that can be seen (FIG. 6A) can be seen smoothly (FIG. 6B). As a result, the image quality can be significantly improved.
[0034]
In the present embodiment, the image information (color, luminance) of the sub-pixel in one pixel often differs from each other as shown by P2 and P4 in FIG. However, since the image information obtained by the present embodiment needs to be displayed on a display device having a finite number of pixels, it is necessary to return the information to pixel information. Therefore, in this embodiment, for example, an average value (weighted average value) of the image information of the sub-pixels within one pixel is obtained, and this is used as the image information of the pixel. In the conventional example, the image information of the pixel at the left eye point and the image information of the pixel at the right eye point are the same. However, according to the present embodiment, these are often different. For example, in FIG. 5, the image information of the pixel P1 and the image information of the pixel P2 are different, and P2 is darker. However, even if the image information of the left eye and right eye point of view differs in pixel units, the same effect can be given to human vision as when they are the same, and almost always adversely affects stereoscopic vision. Absent.
[0035]
In the above, the description has been mainly given assuming that the basic pattern image is configured in pixel units. However, the present invention is not limited to this, and the basic pattern image may be configured in sub-pixel units as shown in FIG. In this case, when obtaining the final image information, a process of averaging the image information of the sub-pixels constituting each pixel is also required for the basic pattern image.
[0036]
(Second embodiment)
The second embodiment is an embodiment in which the formation of a basic pattern image is improved to thereby improve the image quality.
[0037]
When an image is generated by the method of the first embodiment, the completed image is composed of an original area and a non-original area as shown in FIG. Here, the original area is an area in which the basic pattern image is simply repeated. For example, in FIG. 1, when all the depth values of the area of the first modified image are 0, the first modified image is the same as the basic pattern image. Therefore, in this case, the areas of the basic pattern image and the first modified image are both original areas. On the other hand, if a display object such as a cube exists in the area of the second modified image, the second modified image is different from the basic pattern image, and therefore, a part or all of this area becomes a non-original area.
[0038]
When an image in which an original area and a non-original area are mixed is displayed on a display device on which gamma correction has not been correctly performed, a difference occurs in brightness between the original area and the non-original area, thereby deteriorating image quality. Occurs. The reason will be described below.
[0039]
In a display device such as a CRT, the display luminance of a pixel on a screen is not proportional to an input value to the pixel and has a characteristic as shown in FIG. Thus, normally, as shown in FIG. 9B, the correction processing is performed so that the display luminance of the pixel is proportional to the input value. This correction process is called γ correction. There is an individual difference in the relationship between the input value and the display luminance, and there is a temporal change. Therefore, the γ correction of the display device is not always performed correctly. For example, when the input value / display luminance curve is convex upward (V in FIG. 10A) or convex downward (FIG. 10A). 10 (A) W) can occur. For example, consider the case where the input value / display luminance curve is convex upward. Also, consider a case where the basic pattern image is not configured in sub-pixel units (see FIG. 7) but is configured in pixel units. In this case, in the original area (the basic pattern image and the area of the same image), processing for averaging the image information of the sub-pixels is not performed, and the image information of the pixel remains the original image information. Become. On the other hand, in the non-original area, as described in the first embodiment, a process of averaging the image information of the sub-pixels is performed, and the image information of the pixel becomes a state in which a plurality of pieces of image information are mixed. ing.
[0040]
For example, in the pixel P2 of FIG. 5, if the image information with the lower display luminance is C1 and the image information with the brighter display is C2, there are four subpixels of C1 and 12 subpixels of C2. Therefore, assuming that the image information of the pixel P2 is C3 and the display luminance of the display device with respect to the image information C1, C2, C3 is b1, b2, b3, respectively.
b3 = (4 · b1 + 12 · b2) / 16
It should be. This corresponds to the case where n = 12 and m = 4 in FIG. That is, the display luminance of the display device must be b3 shown in FIG. However, at this time, if the gamma correction of the display device is not performed correctly and the input value / display luminance curve is convex upward (in the case of V), the display luminance of the display device is b4 instead of b3. The display will be bright. Conversely, if the input value / display luminance curve is convex downward (in the case of W), the display will be dark.
[0041]
On the other hand, in the original area, as shown in the basic pattern image of FIG. 5, the same image information is provided in all the pixels, so that the brightness does not change as described above. For example, when there are four C1 pixels and twelve C2 pixels in the original area, the C1 pixel on the display device is displayed at the b1 display luminance, and the C2 pixel is displayed at the b2 display luminance. Therefore, the human eyes perceive that the display luminances b1 and b2 are mixed at a ratio of 4:12, that is, the display luminance of b3 in FIG. 10B. On the other hand, the display luminance output of b4 is output from the pixel P2 in the non-original area, so that the non-original area has a brighter display.
[0042]
In this embodiment, in order to solve the above problem, as shown in FIG. 11, a basic pattern image is generated such that the size of a region where image information is the same is a non-integer multiple of the size of a pixel. . In the example of FIG. 11, the size of the same image information area is 1.25P × 1.25P (P is the length of one side of the pixel), which is a non-integer multiple of the size of the pixel. As a result, the image information is mixed not only in the non-original area but also in the original area, and both are equally affected by the change in display luminance due to the mixing. As a result, there is no difference in display luminance between the original area and the non-original area, and the image quality can be improved.
[0043]
As shown in FIG. 7, if the basic pattern image is formed so that the same image information area always covers the pixel area, the same pixel information area does not necessarily have to be a non-integer multiple of the pixel size.
[0044]
Note that the present invention is not limited to the first and second embodiments, and various modifications can be made within the scope of the present invention.
[0045]
For example, in the above-described embodiment, description has been given mainly of RDS as an example, but the present invention can be applied not only to RDS but also to other stereoscopic image information generating methods such as color field stereograms. Here, in the color field stereogram, a basic pattern image in which a pattern or the like of a predetermined shape is drawn in pixel units is used. At this time, in the present invention, the basic pattern image can be drawn in sub-pixel units. In addition to the RDS and the color field stereogram, a stereoscopic view in which the image information of the right-eye (or left-eye) fixation point is determined based on at least the image information of the left-eye (or right-eye) fixation point and the distance between the fixation points. The present invention can be applied to any image information generating method. In this case, the distance between fixation points is set in sub-pixel units, and the image information of the right eye (or left eye) fixation point obtained in sub-pixel units is converted into image information in pixel units, which is equivalent to the above embodiment. The effect of can be obtained.
[0046]
In the above-described embodiment, a method of calculating an average value when converting image information in sub-pixel units into image information in pixel units is employed. However, the present invention is not limited to this. For example, various methods such as performing γ correction together with these conversions can be adopted.
[0047]
The image information generating method of the present invention has a particularly great effect when the obtained image information is displayed on a display device having a finite number of pixels. However, there are other effective uses. For example, it is used for generating image information by the method of the present invention in a personal computer or the like, or for distributing the generated image information by communication (such as personal computer communication). When image information is generated on a computer, a characteristic such as a bitmap memory for image display is often used. In such a case, due to the properties of the bitmap memory and the like, there is a restriction that the finally generated image information must be in pixel units. In the case of delivering image information by communication or the like, it is necessary to reduce the data amount of the image information, and therefore, there is a restriction that the finally generated image information must be in a predetermined unit, for example, a pixel unit. According to the present invention, a high-quality stereoscopic image can be provided even under such restrictions.
[0048]
Further, according to the present invention, a moving image can be represented by the generated image information. When a moving image is represented by a conventional method, a step-like shape as shown in FIG. 6A may appear or disappear depending on the direction of a display object. According to this, such a situation can be prevented.
[0049]
【The invention's effect】
According to the present invention, since the distance between fixation points can be controlled in sub-pixel units, a sense of depth can be accurately expressed, and a stereoscopic image that generates a virtual image having a smooth inclined shape can be generated. As a result, the quality of the obtained image can be greatly improved. Then, according to the present invention, the image information of the left eye and the right eye point of view in pixel units may be different, but the same effect can be given to human vision as when they are the same, and the stereoscopic effect can be obtained. Has little adverse effect on vision.
[0050]
Further, according to the present invention, even on an image display device or the like in which γ correction has not been correctly performed, it is possible to display an image with uniform brightness as a whole, and it is possible to further improve the quality of an obtained image.
[0051]
[Brief description of the drawings]
FIG. 1 is a diagram showing an example of an image obtained by the present invention.
FIG. 2 is a diagram for describing three-dimensional shape and depth information necessary for generating an image.
FIG. 3 is a diagram for explaining an RDS method.
FIG. 4 is a diagram for explaining a relationship between a pixel and a sub-pixel.
FIG. 5 is a diagram for describing an image information generation method according to the first embodiment.
FIGS. 6A and 6B are diagrams for explaining a problem of the conventional method.
FIG. 7 is a diagram for describing a basic pattern image configured in sub-pixel units.
FIG. 8 is a diagram illustrating an example of an arrangement of an original area and a non-original area.
FIGS. 9A and 9B are diagrams for describing γ correction. FIG.
FIGS. 10A and 10B are diagrams for explaining a problem that occurs when an image is displayed on a display device on which γ correction has not been correctly performed.
FIG. 11 is a diagram illustrating an example of a basic pattern image generated in the second embodiment.
FIGS. 12A and 12B are diagrams for explaining an intersection method and a parallel method.
[Explanation of symbols]
10 balls 12 cubes 14 walls

Claims (11)

Based on one of the left-eye and right-eye gazing point image information and the distance between gazing points determined by the depth information, the other is different from the one and determines the other gazing point image information in a pixel unit that can be stereoscopically viewed. An image information generating method for generating image information,
Obtaining the one point-of-interest image information in sub-pixel units, and the other point-of-interest image information in sub-pixel units based on the inter-point-of-regard distance set in sub-pixel units;
Obtaining image information in pixel units from the other point-of-interest image information in sub-pixel units. An image information generating method for forming a plurality of adjacent modified image information based on basic pattern image information and a distance between gazing points determined by depth information to generate stereoscopically visible pixel-based image information. hand,
Converting the basic pattern image information in pixel units into image information in subpixel units;
Based on at least one of the basic pattern image information in sub-pixel units, the (N-1) th modified image information (N is an integer) in sub-pixel units, and the distance between gazing points set in sub-pixel units, Generating N modified image information;
Obtaining image information in pixel units from each of the obtained modified image information in sub-pixel units. In any of claims 1 or 2,
An image information generating method, wherein in the step of obtaining the pixel-based image information, an average value of image information of sub-pixels included in each pixel is used as the pixel-based image information. In any of claims 2 or 3,
The basic pattern image information is composed of sub-pixel unit image information, and in the step of obtaining the pixel unit basic pattern image information, pixel unit image information is obtained from the sub pixel unit basic pattern image information. Image information generation method. In any one of claims 2 to 4,
An image information generating method, wherein the basic pattern image information is formed by arranging images randomly in pixel units or subpixel units. In any one of claims 2 to 4,
An image information generating method, wherein the basic pattern image information is formed by drawing a pattern of a predetermined shape in pixel units or sub-pixel units. In any one of claims 4 to 6,
An image information generating method, wherein the basic pattern image information is formed such that at least one of the vertical and horizontal dimensions of a region where the image information is the same is a non-integer multiple of the pixel dimension. In any one of claims 4 to 6,
An image information generating method, wherein the basic pattern image information is formed such that a region where the image information is the same always crosses a pixel boundary. An image display method, comprising: displaying image information generated by the image information generation method according to claim 1 on a display device having a finite number of pixels. In claim 9,
An image display method, comprising displaying the image information of a moving image on the display device. Based on one of the left-eye and right-eye gazing point image information and the distance between gazing points determined by the depth information, the other is different from the one and determines the other gazing point image information in a pixel unit that can be stereoscopically viewed. An image display system that generates image information and displays it on a display device having a finite number of pixels,
Means for obtaining the other point-of-interest image information in sub-pixel units, based on the one point-of-interest image information in sub-pixel units and the distance between points of interest set in sub-pixel units,
Means for obtaining image information in pixel units from the other point-of-interest image information in sub-pixel units.

Images