JP3144637B2 - 3D rendering method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a three-dimensional rendering method for rendering an object in three-dimensional image processing.

[0002]

2. Description of the Related Art Conventionally, in synthesizing a three-dimensional image, rendering is performed on a display object in a visual field region in real time or at regular intervals, or whenever the object changes in color, position, shape, or the like. To do. However, if rendering is performed every time for all the display objects in the visual field, the calculation load becomes enormous, and the rendering process takes a long time. Therefore, conventionally, various methods for reducing the calculation load and obtaining a natural three-dimensional image have been proposed.

FIG. 10A is a diagram showing a region in a three-dimensional space to be rendered in a process of synthesizing a three-dimensional image of, for example, a game machine, and FIG.
It is the figure which drawn the area of (a) two-dimensionally. As shown in FIG. 10, in general, three-dimensional rendering is performed only on the area of the truncated pyramid sandwiched between near and far, and a background that is a two-dimensional image is pasted on the area behind it. Thus, the calculation load is reduced without performing the rendering process on all the areas.

Japanese Patent Application Laid-Open No. 6-348860 discloses, for example, in a drive game, an image of a three-dimensional object (a passing building or the like) having a specific unevenness in a region viewed from a predetermined angle is stored as texture information in advance. In addition, an image synthesizing apparatus that forms a display screen by mapping this texture information to a polygon of a three-dimensional object is disclosed. Thus, an object having unevenness, which conventionally required as many as 100 polygons in a three-dimensional area, is represented by a small number of polygons, thereby reducing the hardware load.

In Japanese Patent Application Laid-Open No. Hei 9-167258, a display object (original three-dimensional object) such as a tree arranged in a three-dimensional area is formed as a simple object (planar image), and there is no displacement of the viewpoint. However, there is disclosed an attempt to reduce the calculation load and reduce the processing time by rotating the object so as to face the line of sight and performing composite display.

[0006]

However, FIG.
In the example shown in (1), since the area defined by near and far has a certain size, the hardware load is still large to perform the rendering processing on all the areas in real time. To avoid this, ne
It is conceivable to narrow the ar-far region, but in that case, the space for performing the three-dimensional image processing becomes narrow, and various problems occur.

On the other hand, according to the technique disclosed in the above-mentioned Japanese Patent Application Laid-Open No. 6-348860, a plane (two-dimensional) image of a display object viewed from a predetermined angle is prepared, and the image is displayed in accordance with the distance of the display object. By changing the size of the polygon on which this image is mapped, a three-dimensional perspective is realized.
However, this technique can be suitably used when the viewing angle of the display object is within a predetermined range, such as a roadside building in a driving game, but the player changes the viewing angle while moving. In the case of a simple game, the image cannot be used. For this reason, there is a description that an image viewed from a plurality of angles is prepared so as to be able to cope with a change in the viewpoint, but since this image is a planar (two-dimensional) image, it is particularly close to the display object. In some cases, the unnaturalness does not disappear as compared with a stereoscopic image in which unevenness is displayed using a large number of polygons.

In the technique disclosed in Japanese Patent Application Laid-Open No. 9-167258, when the angle from a display object, which is a simple object, changes, only the direction of the object is changed to perform composite display. The simple display (two-dimensional image) remains even when the subject is close to the target. Therefore, also in this case, particularly when approaching the display object, unnaturalness remains as compared with an image stereoscopically displayed using a large number of polygons.

As described above, according to the prior art, since a planar image is used for a specific object in a three-dimensional area, high-speed rendering is possible. It is only possible to deal with the angle, and even if the target object is close, it is a two-dimensional image. Therefore, when the viewpoint moves close, a real image close to the actual state cannot be obtained.

An object of the present invention is to solve the above-mentioned problems,
It is an object of the present invention to provide a three-dimensional rendering method capable of performing high-speed rendering and obtaining a realistic image close to an actual state when a viewpoint approaches a target.

[0011]

A three-dimensional rendering method according to the present invention is a method for rendering an object in a virtual three-dimensional space, in which a virtual boundary surface is assumed at a predetermined distance from a viewpoint. When an area closer to the viewpoint is a first area and an area farther from the virtual boundary surface than the viewpoint is a second area, (1) rendering is performed on an object in the first area and the second area. And using the image of the second area obtained by this rendering as a background image,
In the subsequent rendering, the rendering is continuously performed only on the object in the first area, and
(2) When an image shift occurs between the background image and the image rendered for the object in the first area and the background image needs to be updated, the object in the second area is updated. And updates the background image.

According to the present invention, if a part of the three-dimensional region that is far from the virtual boundary is rendered, it is used as a background and then used as a background, until the image of the foreground deviates from the virtual boundary. By configuring so that re-rendering is not performed, the number of times of re-rendering for a three-dimensional area beyond the virtual boundary surface is greatly reduced, enabling high-speed rendering. At the same time, a natural image close to the actual state can be obtained by using the continuously rendered image for the object close to the viewpoint.

The present invention also relates to a method for rendering an object in a virtual three-dimensional space, wherein the first virtual boundary surface is closer to the predetermined distance from the viewpoint than the first virtual boundary surface when viewed from the viewpoint. Assuming a second virtual boundary surface at the position, a region closer to the viewpoint with respect to the first virtual boundary surface is a first region, and a region farther from the viewpoint with respect to the second virtual boundary surface is a second region. (1) First area and second area
Rendering is performed on the object within the region, and the image of the second region obtained by this rendering is used as a background image, and subsequent rendering is continuously performed only on the object within the first region. And (2) when an image shift occurs between the background image and the image rendered for the object in the first area and the background image needs to be updated, the second The background image may be updated by rendering the object in the area.

As will be described later in detail, if image processing is performed separately on the near side and the back side of one virtual boundary surface, the first area and the second area can be used, for example, when the viewpoint retreats. There is a case where the blank part is lost without rendering anything. Therefore, if the first area for performing the three-dimensional rendering in real time and the second area for performing the re-rendering in a predetermined case partially overlap in advance, the movement and rotation of the viewpoint within a certain range may occur. However, no blank portion of the image is generated between the first area and the second area, and it is possible to prevent the image from being rendered black without being rendered.

[0015] As a preferred example of the present invention, the updating of the background image due to the shift of the image between the background image and the rendering image in the first area is performed by moving the viewpoint so that the distance from the viewpoint to the virtual boundary surface is 5%. It is configured to be performed when the line of sight changes, or when the line-of-sight direction changes by 5 degrees or more due to line-of-sight movement. According to the simulation by the inventor of the present application, it has been found that a natural image without a sense of incongruity can be obtained by updating the background image in such a case.

Further, as a further embodiment, the updating of the background image due to an image shift between the background image and the rendering image in the first area is performed when a predetermined time has elapsed in the virtual three-dimensional space. It is configured to be performed at the time. In this case, optimal three-dimensional rendering can be performed particularly in a cityscape simulation.

[0017]

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. FIG. 1 (a),
1B is a diagram for explaining the three-dimensional rendering method of the present invention, FIG. 1A is a diagram showing a region in a three-dimensional space to be rendered, and FIG. It is the figure which drawn the area of (a) two-dimensionally. The three-dimensional space to be rendered in the present invention is, as shown in FIG. 1A, a quadrangular pyramid region extending from a viewpoint (eye or camera) 1, and a truncated quadrangular pyramid sandwiched between near and far. Is a three-dimensional area to be rendered.

As shown in FIG. 1, in the present invention, near and fa
The virtual projection plane P is set in the area (three-dimensional area) defined by r, and an area closer to the virtual projection plane P is defined as a first area, and an area farther than the virtual projection plane P is defined as a second area. here,
It is preferable that the virtual projection plane P is set in an arrangement orthogonal to the visual axis direction from the viewpoint 1. Then, the display object 2 located farther from the virtual projection plane P, that is, in the second area,
The image is projected on the virtual projection plane P as a two-dimensional image, and is used as a background image. At the same time, for the display object 3 between the viewpoint 1 and the virtual projection plane P, that is, the display object 3 in the first area, a calculation for displaying in real time the same three-dimensional virtual space as in the related art is performed. Drawing is performed in a virtual space (near) with the space plane P.

FIG. 2 is a flowchart for explaining an example of the two-dimensional rendering method of the present invention. In the rendering method of the present invention, first, three-dimensional rendering is once performed on all objects in the first area and the second area. That is, after rendering the object in the second area, it is stored as a two-dimensional background image (step S1). Then, the object in the first area is rendered using the background image obtained here (step S2).

As an actual example, FIG. 3 shows a rendered image of a background in the second area, FIG. 4 shows a rendered image of a foreground in the first area, and FIG.
5A and 5B respectively show images obtained by combining a rendered image of a background in a region and a rendered image of a foreground in the first region shown in FIG.

Thereafter, rendering is continuously performed on the first area. Here, it is determined whether or not the background image needs to be updated due to the influence of movement or rotation of the viewpoint (step S3). . If it is determined that the background image does not need to be updated (NO), rendering of only the first area is repeated while the background of the second area remains unchanged.

In step S3, when it is necessary to update the background image, that is, when there is a shift between the background image and the re-rendered image in the first area (YES), the second Are re-rendered, and the two-dimensional image is used as a new background image to synthesize the image.

Here, the determination of the update time of the background image due to the image shift between the background image and the rendering image in the first area (FIG. 2: step S3) will be described. 6 (a) to 6 (c) and FIG. 7 are diagrams for explaining an example of a method of determining the update time. 6 (a) to 6 (c)
By moving the viewpoint, two rendering areas, ie, the first
FIG. 6A shows an example in which the image is displaced between the first area and the second area, and FIG. 6B shows an example in which the second area is retracted from the first area. 6C is a plan view showing an example in which the second area has moved forward with respect to the second area, and FIG. 6C is a plan view showing an example in which the second area has moved up, down, left, and right with respect to the first area. FIG. 7 shows a case where a shift occurs between the two rendering areas, that is, the first area and the second area due to the rotation in the viewing direction.

In the present invention, since the rendering in the second area is not normally performed, the rendering in the first area is repeated, and as shown in FIG. 6 and FIG. A shift occurs between the rendered image in the first area and the rendered image in the second area. In such a case, since an area where nothing is rendered occurs and the image is blackened, or the image of one area is hardly unnaturally contained in the image of the other area, the image is not transferred to the second area. Rendering is required.

The basis for re-rendering the second area is that, in the cityscape simulation, updating of the background image due to an image shift between the background image and the rendered image in the first area is performed by moving the viewpoint. This is performed when the distance from to the virtual boundary surface changes by 5% or more, or when the line of sight changes by 5 degrees or more due to line of sight movement. In addition, according to the simulation by the inventor of the present application, it has been found that the deviation between the two regions does not cause a problem up to a displacement of about 5 to 7% in distance and about 5 degrees in angle.

The reference for re-rendering the second area is not limited to the above example. As an example, in the case of a city landscape simulation, an example in which the position of the sun shifts with time in a three-dimensional space will be described with reference to FIG. As shown in FIG. 8A, at time a in the virtual space, since the light of the sun 5 is irradiated at an angle a1, a shadow a2 is formed by the display object 2 in the second area. Thereafter, at time b, the sun 5 moves, so that light is emitted at an angle b1 and the shadow of the object 2 should be b2. In this case, re-rendering of the second area is performed even if the viewpoint does not move at all. Alternatively, rendering may be performed at regular time intervals or whenever there is a change in the color, position, shape, or the like of the display object 2.

By drawing all the display objects in the three-dimensional virtual space in this way, the display object in the second area uses a two-dimensional background image to dramatically reduce the hardware load. In addition, a three-dimensional rendering image close to the actual video can be obtained by performing real-time three-dimensional rendering on the display object in the first area near the viewpoint from the viewpoint.
Therefore, a natural three-dimensional image can be realized as a whole.

FIG. 9 is a diagram for explaining another embodiment of the three-dimensional rendering method of the present invention. As shown in FIG. 9, in the present embodiment, the three-dimensional area is divided into a first area defined by near1-far1 and a second area defined by near2-far2. The first area and the second area are set so as to partially overlap.

The three-dimensional rendering method of the present embodiment also
Rendering is performed in the same manner as in the method of the first embodiment described above. That is, first, three-dimensional rendering is performed on all objects in the first area and the second area. Then, the rendering result of the second area of near2-far2 is set as the background. Then, at the time of the next and subsequent renderings that are repeated at regular time intervals, the rendering of the first area is repeated while the background of the second area remains unchanged. Then, when a deviation of a predetermined amount or more occurs between the background image and the re-rendered image in the first area, it is determined that the rendering result of the second area needs to be updated,
The rendering of the second area is performed again, and the result is set as a background image. By repeating the above steps, real-time rendering can be performed.

If the three-dimensional rendering is performed in this manner, as shown in FIGS. 6 and 7, even if the viewpoint position slightly changes and the two rendering areas are shifted, nothing is rendered. It is possible to avoid a situation where an image is lost in black. The range in which the first region and the second region overlap is set to about 5 to 10%, preferably about 7%, in the distance direction from the viewpoint.

[0031]

As is apparent from the above description, according to the present invention, when the three-dimensional rendering area is divided into the first and second areas and the rendering is performed once for the second area as the background, By using this as a background, re-rendering is not performed until there is a deviation from the first area that renders in real time, so the number of re-renders of distant views is greatly reduced and high-speed rendering is possible In addition, since the background is an image on which three-dimensional rendering has been performed, a realistic image close to the actual state can be obtained.

[Brief description of the drawings]

FIGS. 1A and 1B are diagrams for explaining a three-dimensional rendering method of the present invention.

FIG. 2 is a flowchart illustrating an example of a three-dimensional rendering method according to the present invention.

FIG. 3 is a diagram illustrating an example of a rendered image of a background in a second area.

FIG. 4 is a diagram illustrating an example of a foreground rendered image in a first area.

FIG. 5 is a diagram showing an example of an entire rendering image obtained by combining the rendering images shown in FIGS. 3 and 4;

FIGS. 6A to 6C are diagrams illustrating an example of a method of determining a background image update time.

FIG. 7 is a diagram for explaining another example of a method of determining a background image update time.

FIG. 8 is a diagram for explaining another example of a method of determining a background image update time.

FIG. 9 is a diagram for explaining another embodiment of the three-dimensional rendering method of the present invention.

FIG. 10 is a diagram for explaining a conventional three-dimensional rendering method.

[Explanation of symbols]

 1 viewpoint 2, 3 display object 5 virtual sun

──────────────────────────────────────────────────続 き Continued on the front page (58) Fields surveyed (Int. Cl. ⁷ , DB name) G06T 15/00 100 G06T 15/70 JICST file (JOIS)

Claims (5)

(57) [Claims] 1. A method for rendering an object in a virtual three-dimensional space, wherein a virtual boundary surface is assumed at a predetermined distance from a viewpoint, and a region closer to the viewpoint with respect to the virtual boundary surface is defined as a first region.
The region on the side farther from the viewpoint with respect to the virtual boundary surface
(1) Rendering is performed on the objects in the first and second areas, and the image of the second area obtained by this rendering is used as a background image. The rendering is continuously performed only on the object in the first area, and (2) the image is shifted between the background image and the image rendered on the object in the first area. A three-dimensional rendering method, characterized in that when the background image needs to be updated, the object in the second area is rendered and the background image is updated. 2. A method for rendering an object in a virtual three-dimensional space, wherein a first virtual boundary surface is located at a predetermined distance from a viewpoint, and a second virtual boundary surface is located at a position closer to the first virtual boundary surface as viewed from the viewpoint. Assuming a virtual boundary surface, a region closer to the viewpoint with respect to the first virtual boundary surface is a first region, and a region farther from the viewpoint with respect to the second virtual boundary surface is a second region. Then, (1) rendering is performed on the objects in the first area and the second area, and the image of the second area obtained by this rendering is used as a background image, and the subsequent rendering is performed using the first area. The rendering is continuously performed only on the object in the area, and (2) an image shift occurs between the background image and the image that has been rendered on the object in the first area. Image needs to be updated In this case, the three-dimensional rendering method is characterized in that the object in the second area is rendered and the background image is updated. 3. An update of a background image due to an image shift between the background image and a rendering image in a first area,
5% distance from viewpoint to virtual boundary surface due to viewpoint movement
The three-dimensional rendering method according to claim 1, wherein the method is performed when the change has occurred. 4. An update of a background image due to an image shift between the background image and a rendering image in a first area,
The three-dimensional rendering method according to any one of claims 1 to 3, wherein the method is performed when a line-of-sight direction changes by 5 degrees or more due to a line-of-sight movement. 5. An update of a background image due to an image shift between the background image and a rendering image in a first area,
The three-dimensional rendering method according to claim 1, wherein the method is performed when a predetermined amount of time has elapsed in the virtual three-dimensional space.