JPH02112079A - Picture generating method - Google Patents

Abstract

PURPOSE:To attain the expression of half shadow by considering a surface light source model in a shadow map method, obtaining a ratio to be irradiated from the light source concerning the respective picture elements of a shadow map and storing the ratio to the shadow map. CONSTITUTION:A picture generating method is composed of an input device 1, a central processing unit 2 to execute the execution of a program or data processing, a graphic display device, which displays a generated picture, and a storage device 4. A shadow map generating program houses the information of an object, which is visible from one point of the surface light source model, and the intensity of a light, with which the object is irradiated, to the shadow map. A computing program obtains the intensity of the light, with which the picture element is irradiated from the surface light source model, concerning the respective picture elements of the shadow map by using the prepared shadow map. Then, the intensity of the light is housed to the shadow map. A picture generating program executes luminance computation and shadow processing by referring the shadow map and considering the light, with which the picture element is irradiated from the surface light source, in the picture generation from a view point. Thus, the expression of the half shadow can be executed at a high speed.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a system that generates a color shaded image from object data that defines the shape of a three-dimensional object, and in particular takes into account a surface light source and includes a penumbra. This invention relates to a computer graphics system that generates realistic images.

[Conventional technology]

Displaying shadows is important because it makes the positional relationship of objects clear and increases the sense of reality. For this reason, various methods for processing trellis shapes have been proposed in computer graphics. In order to generate a realistic image, it is essential to display a soft shadow at the border, that is, a penumbra.

Up until now, computer software has been used as a fast way to display shadows.
Compute, Graph, 12.3 (1978), pp. 270-274.

Vol, 12. No, 3. pp, 27O-274(
Williams' algorithm (shadow map method) discussed in 1978), ) is known. This method consists of two steps: first order;

■ Remove hidden surfaces from the displayed object using the depth buffer method using the light source as a virtual viewpoint. At this time, the brightness is not calculated, and only the depth value (depth) is stored in the buffer.

■ Generate an image from the viewpoint using the depth buffer method. At this time, each pixel is converted to a coordinate system related to the light source, and
Find the visibility from the light source using the buffer depth value found in . If it is invisible from the light source, perform monk shape processing.

A major feature of this algorithm is that it does not choose the processing target. In other words, not only polygons but also free-form surfaces,
Can handle fractal surfaces. In addition, the processing is simple, and the amount of processing when performing monstrosity processing is as low as 1.

It is 1 to 2 times less than when no monk processing is performed. Since a parallel light source or a point light source is assumed, the shadow will be sharp without a penumbra.

However, with this method, it is impossible to display a vague shadow of the boundary, that is, a penumbra.

What are the penumbra display methods that have been proposed so far?

(1) A method of determining in advance the surface where the penumbra 9 wood carving occurs, (2) Extension of the shadow volume method, (3) Distributed ray tracing method (D1 distributed RayTraci)
ng).

There are three.

Method (1) is published in Information Processing Society of Japan Journal, 23, 4 (198
2nd year) As discussed on pages 373 to 380, first find the part where the penumbra and wood carving occur for each polygon. During display processing, the shading is calculated by determining the light intensity irradiated from the line light source in the 5th penumbra area. According to this method, accurate illuminance can be calculated. However, as the scene becomes more complex, the amount of calculation is expected to become enormous.

Method (2) is described in IE, Computer Graphics and Applications, 4.10.
(1984) pages 5 to 12 (IEEE C
G&A, Vol. 4. No, 10°pp, 5-12
(1984), a shadow area (shadow volume) is obtained, and during one display process, it is determined whether each part is within the shadow or not, and the monk shape processing is performed. By approximating a surface light source with a large number of point light sources and applying this method, a penumbra display can be realized. Shadow volumes are generated for polygons, so they are only applicable to polygons.

Method (3) is computer graphics, 18.3
(1984) pp. 137-145 (Comp
ut, Graph,, Vol, 18, No.
3.

Pρ, 137-145 (1984)), it is possible to represent not only penumbra, but also motion blur, depth effects, blurred reflections, and refraction.

In this method, as many rays as necessary are traced.

Appropriate methods of distributing rays have been proposed, but the amount of calculation required is much greater than that of ordinary ray tracing.

As described above, conventional penumbra display methods have had the problem of requiring an enormous amount of calculation.

[Problem that the invention seeks to solve]

The shadow map method can display shadows at high speed, but cannot express penumbra. On the other hand, other shadow display algorithms can express a penumbra, but they have the problem of a huge amount of calculation and slow processing.

An object of the present invention is to provide a high-speed image generation method capable of expressing a penumbra.

[Means for solving the problem] The above purpose is to solve the problem by considering a one-plane light source model in the shadow map method, calculating the proportion of irradiation from the light source for each pixel of the shadow map, and storing it in the shadow map. , achieved.

[Effect]

The shadow map generation program stores information about an object visible from a point on the surface light source model and the intensity of light irradiated to that object in a shadow map, and the illuminance calculation program uses the created shadow map to create a shadow map. The image generation program calculates the intensity of light irradiated to each pixel from the one-plane light source model and stores it in the shadow map.
Luminance is calculated by referring to the shadow map and considering the light from the surface light source that illuminates the pixel.

By performing shading processing, a penumbra display can be realized.

(Example) Figure 1 shows a configuration diagram of this example. 1 is an input device that handles input from the user, 2 is a central processing unit that executes programs and processes data, and 3 is a display unit that displays generated images. Graphic display device.

41 is object definition data that stores object data to be displayed; 42 is an image generation program that generates an image; and 43 is generated image data.

FIG. 2 is an explanatory diagram for simply explaining the penumbra display method. Although this diagram is depicted in two dimensions for simplicity, it is actually three dimensional. In the figure, 101 is a light source model having an area, that is, a surface light source model, 102, 10
3 is the object to be displayed. In this figure, it is assumed that object A is a one-sided light source that casts a shadow on object B, so the shadow will be vague and include a penumbra. In object B.

Part 104 and part 108 are non-shadow parts, 105
The part 107 and the part 107 are penumbra parts that are irradiated with part of the light from the light source, and the part 106 is the umbra part that is not irradiated with any light from the light source.

In order to display a penumbra, it is necessary to find a part (105, 107) that is irradiated with only a part of the light from the light source, and to find the light that is irradiated from the surface light source in that part.

FIG. 3 is an explanatory diagram for explaining the data structure of the shadow map used in this embodiment. In the traditional Williams algorithm, in preprocessing.

The area illuminated by light from a point light source was determined and stored in a depth buffer from the light source, that is, a shadow map.

By using this shadow map, it was possible to determine whether or not an arbitrary point in three-dimensional space is within the shadow.

In contrast, in the present invention, the shadow map is expanded,
Stores the intensity of light emitted from a surface light source. However, the shadow map is a single point within the surface light source (the origin of the light source coordinate system).
Created based on. During shadow calculation, the penumbra is expressed using the light intensity value from the surface light source stored in this shadow map.

The structure of the expanded shadow map is shown in FIG. In the figure, 201 is a shadow map. For each pixel of the shadow map 201, as shown at 202, in addition to the depth value Z, one figure number Id and the irradiation rate R are stored. The irradiation rate R is the rate at which a point is irradiated with light emitted from a light source. This is 0 for the umbra, O~1 for the penumbra, and 1 for other bright areas. Depth value Z
and the figure number rd are determined in step 1, and the irradiation rate R is determined in step 2.

A shadow map is created centered on the origin 0 of the light source coordinate system. Since a surface light source is assumed, even parts that are not visible from the origin O may be irradiated with light. In order to maintain the irradiation rate of such areas, the shadow map is multi-layered. this is.

Achieved with pointers.

FIG. 4 is a flow chart showing an outline of the operation of this embodiment. First, objects visible from the vertices of the light source coordinate system (for example, the center of a surface light source) are determined using a known depth buffer method and stored in a shadow map.

Note that the configuration of the shadow map will be described in detail later.

The shadow map includes not only the object in the foreground.

Information about the second and subsequent objects is stored in the shadow map.

For this purpose, a multi-layered buffer is prepared.

The figure number and depth value Z (depth) are stored in the shadow map (302).

Next, for each pixel of the shadow map, out of the light emitted from the surface light source, the light that is not blocked by other pixels is
The rate at which pixels are reached is determined (303). Hereinafter, this ratio will be referred to as "irradiation rate".

Next, an image is generated from the viewpoint. At this time, each pixel is converted to the light source coordinate system, and the shadow map obtained in step 303 is referred to using the figure number Id and the depth value Z to obtain the light irradiation rate at that point. Shadow calculation is performed with reference to this irradiation rate (304). With the above steps, it is possible to display an image with a penumbra.

302 is the same as Williams' algorithm except that it has multiple layers of buffers. 303 constitutes a feature of this method, and will be explained in detail below. 304 is the same as the Williams algorithm except that the light irradiation rate is used in the shape processing.

Next, the method of calculating the irradiation rate (303) will be explained using FIG. In FIG. 5, 501 is a surface light source model, point 0 is the origin of the light source coordinate system, and point C1 is the end point of the surface light source. 502 and 503 are object data to be displayed. In this figure, object 502 casts a shadow on object 503. The object 502 is expanded into a plurality of pixels on the shadow map. 504 is a pixel (A
). This pixel A creates an area where light from the single-surface light source model 501 is blocked. This area is a triangle PAQ in the figure. however.

Point P is a point sufficiently far from the light source on the straight line DA, and point Q is sufficiently far from the light source on the straight line CA.

Pixels existing within this triangle PAQ are determined.

For this reason, the virtual surface light source 505 is rotated by 180 # around two axes and translated in parallel to the position of pixel A.
Assume (EF). A cone-shaped region RO8 composed of the virtual surface light source 505 and the origin 0 includes a triangle PAQ. Therefore, find the pixels in the area RO8,
It is determined whether the pixel exists within the area PAQ.

The pixels within the region RO8 can be easily determined by associating virtual surface light sources with pixels of the shadow map.

For pixel B existing in area PAQ, the rate at which light from the one-sided light source is blocked by pixel A is determined and subtracted from the illumination rate of that pixel.

The ratio of light shielding by pixel A to pixel B is determined as follows. Pixel A is projected onto the light source side with pixel B as the center. At this time, pixel A is treated as having a size.

The ratio of the projected area within the surface light source is the required shielding rate.

First, the illumination rate of all pixels of the shadow map is initialized to 1, and the above process is executed for all the surface elements to find the illumination rate of each pixel.

FIG. 6 is a flowchart of calculation of irradiation rate (303).

First, the illumination rate of all pixels in the shadow map is initialized to 1 (601). , the following process is repeated for all pixels A in the shadow map (602). A virtual surface light source 50 in which the surface light source model is rotated by 180 degrees around the Z axis and translated in parallel to the position of pixel A.
5 (EF) (603). Next, pixels of the virtual surface light source model are expanded around the origin O (604).
, the following processing is performed for all the developed pixels B (605). That is, the ratio at which pixel B is shielded from light from a single surface light source by pixel A is determined (606), and subtracted from the illumination rate of that surface element B (607). With the above steps, the irradiation rate of each pixel is determined.

In the above embodiment, as a method for determining the irradiation rate, processing was performed mainly on pixels that cast shadows. In addition to this method, a method that processes mainly the light source can be considered.

In a method that processes mainly light sources, a surface light source is approximated by a large number of point light sources. For each point light source, create a shadow map for that point light source. This is achieved by inverse perspective transformation of each pixel of the original shadow map to the world coordinate system, and then perspective transformation of the sampled point light source. Through this conversion, pixels that have changed inside and outside the shadow are found, and the illumination rate is changed for those pixels. The irradiation rate can be calculated using the above steps.

Since the method of the present invention uses the shadow map method, it can be applied to all objects whose hidden surfaces can be removed using the depth buffer method. Furthermore, it does not depend on the number of objects, and the amount of processing is small.

〔Effect of the invention〕

According to the present invention, penumbra can be expressed at high speed.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention, FIG. 2 is an explanatory diagram for explaining a penumbra display model, and FIG. 3 is an explanatory diagram for explaining the data structure of a shadow map. FIG. 4 is a flowchart showing an overview of the process, FIG. 5 is an explanatory diagram showing a method for determining the irradiation rate, and FIG. 6 is a flowchart of the process for determining the irradiation rate. Explanation of symbols 41...Object definition data, 101...Surface light source model, 201...Shadow map. Fire 3 on the 4th eye

Claims (1)

[Claims] In a system that generates a two-dimensional color shaded image from graphic data that defines the calculation of a three-dimensional object, a light source model having an area, that is, a surface light source model is assumed as the light source model, and the surface light source model is An object that has a shadow map to store information about an object visible from a single point and the intensity of light irradiated to that object, and is displayed using the depth buffer method using one point of a surface light source model as a virtual viewpoint. Remove hidden surfaces, store the depth value at each pixel in a Jadow map, and use the created shadow map to find the intensity of light irradiated to that pixel from the surface light source model for each pixel in the shadow map. , is stored in a shadow map, and in image generation from a viewpoint, brightness calculation and shading processing are performed by referring to the shadow map and considering light from a surface light source that illuminates the pixel. Method.