JP2771886B2 - Two-dimensional image generation method based on binocular vision - Google Patents

Description

Description: BACKGROUND OF THE INVENTION The present invention relates to a method for generating, by computer graphics, a two-dimensional image simulating an image of one or a plurality of objects observed by a human with binocular vision. About.

(Prior Art) As a most advanced display method for realistically representing an object, a method of displaying a three-dimensional image has been conventionally studied. As a typical method of displaying a three-dimensional image, a stereoscopic image display method using binocular parallax is used.

In this method, images viewed from the positions (viewpoints) of each of the left and right eyes are independently created and independently presented to the left and right eyes. Humans combine these two images in their heads,
An image with a three-dimensional effect can be perceived. Therefore, this method can provide a human with a three-dimensional realistic image. However, it does not give any information about what kind of image a human is combining based on the independent images presented to both eyes.

Also, in order to view a stereoscopic image using the present method, there are several methods, for example, a method using a twin-lens viewer for viewing an old stereo photograph, and time multiplexing on a CRT (television screen). Although a method of displaying and viewing using polarized glasses has been devised, any additional method requires some additional device in order for a human to observe a stereoscopic image. Problems have been pointed out, such as that the relationship between convergence and accommodation of both eyes is different from the case where a real image is viewed in the natural world, which may cause fatigue.

On the other hand, in order to create and display a general two-dimensional image using computer graphics, a method has been proposed in which one viewpoint is determined and an image equivalent to a camera placed at a position serving as the viewpoint is generated. The two-dimensional image generated in this manner is equivalent to an image when a person closes one eye and directly observes a scene with the other eye. Therefore, even if the image is generated by placing the viewpoint at the center of both eyes, the scene observed by humans using both eyes is (2)
This is different (even if the depth information is lost by the dimensioning), which is different from the two-dimensional image actually sensed by humans.

Accordingly, it is impossible to realistically generate a two-dimensional image simulating a scene actually perceived by using both eyes using computer graphics using any of the above methods. .

An example of creating and displaying the two-dimensional image using computer graphics will be described below.

FIG. 2 is a bird's-eye view showing the arrangement of the objects, and will be described on the assumption that a cylinder 1 having characters drawn on its side surface and a partition 2 is placed on its back surface. The letters drawn on the side of the cylinder 1 are six letters of alphabets A, B... F as shown in the side view of the cylinder shown in FIG. 3, and the marks 3 and 4 on the screen 2 are the diameters of the cylinder. It is assumed that it is drawn with a width equal to D.

In the related art, one viewpoint is determined, and for the purpose of faithfully reproducing a two-dimensional image viewed from that viewpoint, for example, a method such as ray tracing [Hariton (Koriyama translation): Computer graphic by algorithm and program , McGraw-Hill Publishing, 1988].

That is, as shown in FIG. 4, the range of the area visible in the front of the cylinder 1 (hereinafter, referred to as a visible surface) is θ _{1 as} viewed from the viewpoint 5.
≦ 180 degrees, and the range of the visible surface decreases as the viewpoint 5 approaches the cylinder 1. At this time, the characters on the side surfaces of the cylinder 1 are visible surfaces in which only the alphabet "C, D" on the front (0 degree) can be seen, as is clear from FIG. The marks 3 and 4 drawn on the partition 2 are areas that are hidden behind the shadow of the cylinder 1 and cannot be seen (hereinafter, referred to as hidden surfaces). That is, the marks 3 and 4 located inside the positions χ ₁ and χ ₂ of the screen 2 on the extension of the lines of sight 6 and 7 tangent to the viewpoint 5 and the cylinder 1 are shadow surfaces of the viewpoint 5.

FIG. 5 is a block diagram of the apparatus configuration of the method shown in FIG. 4, and in the figure, reference numeral 8 denotes an apparatus for specifying or specifying information such as the position, shape, and texture of an object (referred to as an object position specifying apparatus). Reference numeral 9 denotes a device for determining whether the information of the device 8 is a visible surface or a hidden surface (referred to as an area search device). Ten
Is a device for generating a two-dimensional image on the visible surface determined by the device 9 (referred to as an image output device), and 11 is a device for displaying a two-dimensional image such as a CRT or the like (referred to as a display).
It is.

In the above-described conventional example, the viewpoint 5 is one. On the other hand, when a human looks at the cylinder 1 in a state of binocular vision, the state shown in FIG. 4 is as shown in FIG. In the figure,
51 left perspective, 52 represents the right eye viewpoint, visible surfaces on screen 2 in an extension of the line of sight 12, 13 is a tangent to the cylinder 1 as seen from the left eye viewpoint 51 (region) respectively chi _4, chi ₆ , As well as the right eye perspective
Assuming that the visible surfaces (regions) on the screen 2 which are on the extension of the lines of sight 14 and 15 which are tangents to the cylinder 1 as viewed from 52 are χ ₃ and _{５ 5} , respectively, the visible surfaces (regions) for binocular vision Is located outside of χ ₄ and _{５ 5} and marks 3 and 4 are included in the visible surface.

That is, when the pupil distance (the distance between both eyes) W is wider than the diameter of the cylinder 1, the mark 3 drawn on the screen 2
Enters the field of view of the left eye and mark 4 enters the field of view of the right eye. Then, in a normal state where the field battle is not intense, the human perceives both the marks 3 and 4 in a natural state without being hidden by the shadow of the cylinder 1. Similarly, the area on the side of the cylinder 1 that is synthesized and perceived by the left and right eyes is θ ₂ ≦ 180 degrees, and the value of θ ₂ increases as the distance between the viewpoint and the object decreases. At this time, “BCDE” is visible (including a part of the characters) on the side of the cylinder 1.

Next, FIG. 7 shows an example of an experimental figure obtained by measuring human perceptual characteristics with respect to a figure already drawn on a two-dimensional plane.

In the figure, the distance between marks 16 and 17 is l ₁ , marks 18 and 19
Figure 8 and Figure 7 show graphs of subjective evaluation results comparing psychological lengths of l ₁ and l ₂ when the distance of l is l ₂ (where D <l ₁ , l ₂ ). It is shown in FIG.

In both figures, the average score MOS on the vertical axis, the horizontal axis represents the l ₂ / l _1, The evaluation categories in the table below.

That is, in the subjective evaluation result of FIG. 8 showing a large experimental figure (D = 4.8 cm), when l ₁ and l ₂ are physically substantially equal (l ₂ / l ₁ = 1.0), the psychologically equal size is obtained. I feel it is. On the other hand, a small experimental graphic (D = 1.1 cm) Figure 9 subjective evaluation results of which showed the, l ₂ than l ₁ is the physical 8%
It can be seen that the user feels that the size is psychologically equal in a state where the size is increased (point a crossing the width). These results show that, even if geometrically the same, the subjective length changes when the overall size is different, and the two eyes described earlier with reference to FIG. It is thought to be related to the visual effect.

(Problems to be Solved by the Invention) As described above, the two-dimensional image observed from the single viewpoint 5 in FIG. 4 is observed in the binocular fusion state of the solid viewpoints 51 and 52 in FIG. The two-dimensional image is clearly different from the scene perceived by humans. In other words, on the visible surface (area), when one viewpoint 5 is at the viewpoint 5, the marks 3 and 4 are on the hidden surface, while at the binocular viewpoints 51 and 52, the marks 3 and 4 are included in the visible surface. In other words, the conventional method of generating a two-dimensional image by defining one viewpoint has a problem that a two-dimensional image simulating an image perceived by a human cannot be generated.

(Object of the Invention) It is an object of the present invention to solve the above-mentioned problems of the prior art and to provide a method for generating a two-dimensional image simulating an image viewed by both eyes with human eyes.

(Means for Solving the Problems) In order to solve the above problems and achieve the object, the present invention relates to a method for generating a two-dimensional image using a computer, wherein one or two or more pixels included in an image region to be generated are included. Means for designating or specifying information such as the position, shape, and texture of a plurality of objects; and two different viewpoints, and a visible surface and a shadow visible on the table of the object when the object is viewed from each viewpoint. Area searching means for determining a hidden surface hidden in the image, means for synthesizing a visible area of the image viewed from the two viewpoints obtained by the area searching means, and a synthetic visible surface obtained by the area synthesizing means. Means for correcting information such as the position, shape, and texture of the object so as to be equivalent to that perceived by binocular vision by a human; and a two-dimensional image based on the correction result by the correction means. Means for generating and outputting.

(Operation) The present invention combines the visible planes of an image viewed from two viewpoints in the same manner as a human being by the above means, and based on the obtained combined visible plane, positions, shapes and textures of objects (elements having certain rules). (See Data and Image Communication Glossary)) to create a realistic two-dimensional image by correcting information such as that perceived by humans with binocular vision. can do.

(Embodiment) FIG. 1 is a block diagram showing an apparatus for implementing an embodiment of the present invention. In the figure, an object position specifying device 8, an area searching device 9, an image output device 10 and a display 11 have substantially the same functions as those shown in FIG. The configuration is such that a device 21 is added.

Next, the operation function of each of the above devices will be described. The object position specifying device 8 stores information such as the positional relationship, the shape, and the texture of the object as in the related art. The area search device 9 calculates information on the visible surface when the viewpoints 51 and 52 are placed at the positions of both eyes, unlike the related art, and outputs the information to the area synthesis device 20. The area synthesizing device 20 synthesizes visible surfaces viewed from two viewpoints into one. The simplest method is to make the area obtained as the logical sum of the two a synthetic visible surface. In the case of the example shown in FIG. 6, the side surface of the cylinder 1 is a synthetic visible surface that is within the range of θ ₂ , and the screen 2 is a synthetic visible surface outside the regions χ ₄ and _{５ 5.} .

Next, the object position correcting device 21 generates the object information stored in the object position specifying device 8 such that the combined visible surface specified by the region combining device 9 becomes a visible surface when viewed from a single viewpoint. Is corrected, and the result is output to the image output device 10. That is, in the case of the character of the side surface of the cylinder 1, a character in the range of theta ₂ shown in the Figure 6 to theta ₁ region shown in Figure 4 is intended to be mapped to Osamuru so, mark 3 and 4位置₁ and χ
_The composite visible surface is obtained by performing correction so that the composite visible surface is located outside the area ₂ . Based on the information of the object corrected in this way, the image output device 10 generates a two-dimensional image similar to the conventional one, thereby generating a two-dimensional image based on binocular vision.

In the present description, it is assumed that the object information has been input to the object position specifying device 8 in advance, but the object information may be extracted from an image actually captured using a camera or the like. In addition, when obtaining the composite visible surface in the area synthesizing device 9, not only the logical sum is obtained, but also the area may be smaller than the area obtained by the logical sum in some cases. Further, in the object position correction device 21, it is of course possible to correct the size and shape of the object.

(Effect of the Invention) As described above, according to the method of the present invention, it is possible to generate a two-dimensional image simulating an image viewed by both eyes with human eyes, and to obtain a more realistic two-dimensional image. There is.

The method of the present invention is particularly useful for the conventional method, in the case where the binocular fusion image is an image for a short distance and a small object that is significantly different from the image of a single eye, and the object in such a situation. The sense of distance and size can be improved, and natural depiction can be performed. Therefore, it is effective for generating a two-dimensional image such as a product sample requiring strict realism.

Furthermore, even in grasping the entire image of a long-distance object, for example, an urban development or a completed building, a car, etc., the image obtained by a single eye and the binocular fusion image are slightly different. If so, the effect of the present invention is exhibited.

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of an apparatus for realizing the method of the present invention, and FIGS. 2 to 9 are explanatory views for creating and displaying a conventional two-dimensional image using computer graphics. Is a bird's-eye view showing the arrangement of objects, FIG. 3 is a side elevational view of the cylinder 1 in FIG. 2, FIG. 4 is a view showing how to capture an image from one viewpoint, and FIG. 5 is a view in FIG. FIG. 6 is a block diagram of the apparatus configuration of the method, FIG. 6 is a diagram showing how to capture an image from two viewpoints of a human, FIG. 7 is a diagram showing an example of an experimental figure measured for human perceptual characteristics, FIG. FIG. 9 is a graph showing a subjective evaluation result graph obtained by comparing the psychological lengths l ₁ and l ₂ of the experimental figure of FIG. 1 ... Cylinder, 2 ... Partition, 3,4,16-19 ... Mark, 5 ...
... one viewpoint, 6, 7, 12-15 ... line of sight, 8 ... object position specifying device, 9 ... region searching device, 10 ... image output device, 11
... Display, 20 area synthesis device, 21 object position correction device.

Claims (1)

(57) [Claims] 1. A method for generating a two-dimensional image using a computer, comprising: means for designating or specifying information such as the position, shape, and texture of one or more objects contained in an image area to be generated. And set two different perspectives,
Area search means for determining a visible surface and a hidden surface hidden by a shadow when the object is viewed from the respective viewpoints,
Means for synthesizing the visible surface area of the image viewed from the two viewpoints obtained by the area searching means, and information such as the position, shape and texture of the object based on the synthesized visible surface obtained by the area synthesizing means And a means for generating and outputting a two-dimensional image based on the result of the correction by the correction means. Two-dimensional image generation method based on