JP2003115055A - Image generator - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image generator capable of appropriately and efficiently switching and setting a rendering method of each object and effectively reducing calculation loads of a computer further by dynamically switching processing contents of clipping and shading to each object corresponding to an angle range for which a direction of the object under consideration viewing from a view point in a three-dimensional virtual space is a reference or a space area based on a position of the object under consideration. SOLUTION: The image generator which generates a two-dimensional image indicating a scene of the three-dimensional virtual space on the basis of the object set in the space, a light source and the view point is provided with a modeling means for preparing model data of the object, a coordinate transformation means for transforming the model data to the coordinate system of the view point, a projection conversion means for projecting the model data by the view point coordinate system to a projection surface, and a rendering means for switching the rendering method of each object corresponding to the angle range in the direction of the object under consideration from the view point as the reference or the space area based on the position of the object under consideration.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image generating apparatus for switching a rendering method for each object based on an angular range and a spatial region with respect to a target object in a three-dimensional space.

[0002]

2. Description of the Related Art A conventional image generation system, particularly a three-dimensional computer graphics system (three-dimensional C) for generating a two-dimensional image showing a scene in a three-dimensional virtual space.
In the G system), normally, one two-dimensional image is generated by taking the processing steps shown in FIG.

These are outlined below. In general, the processing steps 202 to 205 in the figure are collectively called a rendering process.

Modeling Modeling is the work of creating in a computer the data such as the shape, color and surface properties of an object to be represented in an image. For example, if you are making an image of a human face, create data such as what the surface shape is, what part of the face has what color, and what the light reflectance is. Then, store it in the computer in a format that can be used for later rendering. Such a collection of data is called an object model. For example, when performing shape modeling of a cube as shown in FIG. 1, first consider a modeling coordinate system with one vertex having a cube as the origin as shown in FIG. Then, the coordinate data of the eight vertices of the cube in this coordinate system and the surface loop data are determined as shown in FIG. 4, for example.
The set of coordinate data and the surface loop data thus obtained becomes the model data of the object.

Rendering Rendering refers to creating an image according to the appearance of an object when it is viewed from a certain position after the model is created. Therefore, in order to perform rendering, it is necessary to consider not only the model but also the viewing position (called a viewpoint or a camera) and lighting conditions. When the rendering work is subdivided, it is generally divided into projection conversion, hidden surface removal, shading, and texture mapping.

[Projection Transform] Projection transform is, when viewed from the viewpoint position, for various coordinate values representing a model.
It is to calculate the position on the screen where it looks and convert it to the coordinates on the screen. FIG. 5 shows 4 for projection conversion.
It is a figure showing two coordinate systems.

Before the projection conversion, it is necessary to convert the model coordinates into data in the viewpoint coordinate system. For that purpose, first, for example, each component of the object position in the world coordinate system is added to each component of each vertex coordinate of the model in the modeling coordinate system to obtain each vertex coordinate of the object in the world coordinate system. Next, each vertex coordinate of the object represented in this world coordinate system is converted into data in the viewpoint coordinate system.
Usually, a homogeneous coordinate system is used to represent this coordinate transformation.
The homogeneous coordinate system is a system in which parallel movement as well as rotation scaling can be handled by the same calculation formula. A vector in which a W (usually 1) component is added in addition to the XYZ three components is used, and a 4 × 4 matrix is used for conversion between the vectors. The vector before conversion is (xyz) and the vector after conversion is (x'y '
z ′), the coordinate transformation using the homogeneous coordinate system is expressed as follows using the coordinate transformation matrix.

[0008]

Here, aij is a component representing rotation and enlargement / reduction, and t _x , t _y and t _z are components representing parallel movement.

After the model data in the viewpoint coordinate system is obtained, projection conversion is next performed. To do the projection transformation,
Set one projection plane in space. For simplicity, in FIG. 5, the projection plane is set to the near clipping plane (described later). There are roughly two types of projection conversion, parallel projection and perspective projection. Here, the point in the viewpoint coordinate system is (x, y,
z), and the coordinates after perspective projection conversion are (x'y ') (projection plane coordinate system).

First, parallel projection refers to the case where the viewpoint is at infinity, and is expressed by the following conversion formula.

[0012]

Perspective projection is an operation of projecting a point in space onto a projection surface when viewed from a viewpoint at a finite distance, and is expressed by the following conversion formula.

[0014]

Here, d is the distance from the viewpoint (camera) to the projection plane (near clipping plane).

By using such a projection conversion formula, an operation for obtaining the intersection of the viewpoint and the projection surface is performed for each point of the three-dimensional shape of the object existing in the space, so that the projection is performed on the projection surface as shown in FIG. Get the figure. The following clipping process is executed in the process of projection conversion.

[Clipping] When drawing everything in front of the viewpoint during projection conversion, it may take unnecessary calculation time. Therefore, it is also necessary to determine the drawing area. This drawing area is called a view volume (view space), and the surface closest to the viewpoint (camera) within the boundary of the view volume is called the near clipping surface and the far surface is called the far clipping surface. Clipping means selecting the figure data inside this view volume, deleting the figure data outside, and if there is a figure that intersects the view volume frame, find the coordinates of the intersection and only figure information inside the view volume. Refers to the method of taking out.

For example, as shown in FIGS.
When two planes (plane 1 and plane 2) are considered in the dimensional space, all three vertices that compose plane 1 are outside the view volume, so this plane is quickly excluded from the drawing target (clipping). To be done. Further, since the surface 2 intersects the view volume, the intersection coordinates (C1, C2) are obtained, and then the projection is performed on the projection surface (however, at each vertex after projection, the lighting value or the Z buffer method ( Since it is necessary to hold the Z value used in (described later), the coordinates of C1 and C2 need to be obtained in a three-dimensional space).

In this way, by selecting only the figures to be displayed in the area, the amount of data can be reduced and the intended figure can be displayed in a short time without expressing an extra object.

The polygon converted into the screen coordinate system is further converted into a column of pixels of a two-dimensional image.
This process is converted by scan conversion (Scan Converter
on), or rasterize (Rasterize). The basic idea of scan conversion is to consider the points (sample points) on the screen (projection plane) that correspond to the pixels (usually the center) of the final output image, and then consider the sample points contained inside the polygon. Seek out,
The pixel corresponding to it is filled with the luminance value at that point (obtained by the shading process described below).

This scan conversion (rasterization)
In the process of hidden surface removal, clipping, shading,
For example, texture mapping is performed. Each of these processes will be described below.

[Hidden Surface Elimination] By the hidden surface elimination, which part of the model can be seen from the current viewpoint position,
Determine which part is invisible. Hidden surface removal involves perspective transformation of the shape model when the 3D shape represented by the surface model or solid model is colored, and the hidden surface is hidden behind a 3D shape close to the line-of-sight direction. It refers to the function that deletes the back side surface and displays only the actual surface or part of the surface. Several techniques have been developed for hidden surface removal.

First, it is Z to perform depth comparison in plane units.
It is a sorting method (depth sorting method). The Z sort method is
It is the simplest and fastest method, and it has been hardwareized from early on, but in many cases it is not possible to sort easily because polygons are contiguous or the order of polygons is not uniquely determined <, incorrect display It also has the drawback of being damaged.

Next, the Z-buffer method processes the front-rear relationship for each pixel expressing a polygon, and can accurately perform hidden surface removal. Disadvantages are the relatively large memory requirement for the Z-buffer over all pixels and the inability to perform transparency and anti-aliasing.

Other hidden surface elimination algorithms include:
Although a scan line method or the like that processes the front-back relation for each scan unit of the screen can be mentioned, the description thereof is omitted here.

[Shading] Shading (Sh
is a technique for realistically displaying a shape model in a computer. It determines the surface color and brightness of an object by calculating the shadow taking into consideration the line of sight, light source position, surface inclination, etc. This is a technique that is generally called shading.

The shading model is the basis for performing the shading process. The shading model is a model of a physical phenomenon in which when light is projected from a light source onto a certain substance, the surface of the substance shines, part of the substance is reflected, and the rest is refracted and transmitted. When performing the shading described below, this phenomenon is simplified and modeled to determine the brightness on the object surface. Generally, when a light source at infinity (parallel rays), the brightness L of a certain point P on the object is diffused reflected light (Diffus).
e reflection: Ld), specular reflection light (Sp
It is often determined as the sum of four elements, namely, the optical reflection (Ls), the transmitted light (Transparent Light: Lb), and the ambient light (Ambient Light: Le) (see FIG. 8).

L = Ld + Ls + Lb + Le (Equation-6) When shading a polygon (a surface forming an object), the simplest and fastest method is to calculate the shading only once for each polygon, and set the entire polygon to that color. It is a method of closing. This method is called constant shading (or flat shading)
The brightness value calculated using the normal vector of each polygon and (Equation-6) is set to all the pixels inside the polygon.

However, since a curved surface cannot be expressed smoothly with this, a method called smooth shading was developed. There are two ways to do this. One is glow shading (Goura shading) that obtains the luminance of pixels on each scan line by linearly interpolating the luminance of each vertex.
ud Shading) (Fig. 10
(A)). In glow shading, the brightness of all pixel points inside the triangle is determined from the brightness 1 ₁ , 1 ₂ , 1 ₃ at each vertex of the triangle using a linear interpolation formula. First, the intersection l _a of one scan line and a triangle in Fig, l _b is given by the following equation.

[0030]

From this, the brightness l _p of the point P on the scan line is obtained by the following equation.

[0032]

However, the brightness of each vertex, for example, l _1, is obtained as follows. First, as shown in FIG. 9, the orientations of the surfaces forming the vertices are obtained, and the average value N is calculated by the following equation.
_{Find v} and find the orientation at that point.

[0034] _{N v = (N 1 + N} 2 + N 3 + N 4) / 4 ( wherein -10) However, N _i is the normal direction of each triangle. Next, the brightness l ₁ is obtained by using (Equation-6). As a result, the smoothly changing luminance of each pixel point is obtained, and the smooth shading processing is performed. This method can perform high-speed calculation and is often used for real-time rendering, but it has a problem that highlights do not appear clearly.

Then, the method considered next is a method called Phong Shading, in which the normal direction of each element on the scan line is obtained from the interpolation of the direction of the normal direction of each vertex, and (Equation According to -6), the brightness at that point is obtained. That is, the direction V of the normal line at the intersection of the scan line and the triangle
_a, calculated by linear interpolation from claim V _1, V _2, V ₃ towards the three vertices V _b, further obtains the normal direction V _p of a point on the scan line in a linear interpolation of the V _a and V _b (FIG.10 (b)). Especially in von shading, highlights such as metal surfaces are expressed more realistically, but since the brightness calculation is performed for all normal vectors generated inside the polygon, there is the disadvantage that the amount of calculation is larger than in glow shading. is there.

[Texture Mapping] The texture that can be expressed by manipulating the four parameters of shading (specular reflection light, ambient light, diffuse reflection light, transmitted light) is limited to metal, plastic, gypsum and the like. A variety of materials that are more familiar to you, such as wood, cloth, and marble,
Texture mapping is used for things that cannot be expressed without a “pattern”. FIG. 11 shows a method using perspective coordination (projection coordination). This is a method of mapping the map data from a certain point so that it is projected onto the object.

Specifically, the texture mapping is realized by inserting a pixel reference process for the texture image in the above shading process. The pixel to be referenced is assigned to each vertex of the polygon by the two-dimensional coordinates (UV coordinates) of the original texture image, and the UV coordinates are interpolated during the rasterization process to determine the pixel to be referenced in the original texture image. At this time, if necessary, the reference pixel is subjected to the above-described shading processing to determine the value of one pixel inside the polygon. Pixel reference methods include a method of simply extracting one pixel of the original texture image (point sampling) and a method of referring to a plurality of pixels and averaging them to determine one pixel (filtering). In point sampling, when the area of the original texture image is smaller than the area of the polygon to be rendered, one pixel of the texture image is enlarged and discontinuous boundaries of pixels appear to roughen the image. By performing filtering, discontinuity of pixels can be removed and a smooth image can be obtained. As described above, when accurate interpolation is performed in the UV coordinate complementing method and filtering is performed on the pixel reference, the amount of calculation at the time of rasterization increases.

The general image generation method in the conventional three-dimensional CG system has been outlined above, but such a process requires a relatively large amount of calculation, and the computer to be realized needs to have a certain calculation speed. Has been said.

Until about 7 to 8 years ago, most of the systems introduced with the above-mentioned three-dimensional CG are for business use.
It was only available on high-end computers called graphics workstations (GWS). Therefore, the users who can use it were also limited. However, the situation has changed dramatically over the last few years. With the rapid progress in computer technology, inexpensive and high-speed devices such as CPUs (Central Processing Units) have been mass-produced, so that the number of 3D CG software for personal computers has increased rapidly, while 3 A high-performance home-use game machine using dimensional CG has also appeared. And since all of them are available at relatively low prices, 3
The hardware / software using the three-dimensional CG has become widespread among general users.

[0040]

However, such a conventional three-dimensional CG system has the following problems.

In the conventional general three-dimensional CG system,
When a plurality of objects are set in a three-dimensional virtual space, usually, each object is uniformly rendered by high-quality clipping processing or shading processing to generate a two-dimensional image. In this case, needless to say, the computational load of the computer becomes large, and when rendering and displaying a plurality of models having a complicated shape in real time, the image may not move smoothly.

To avoid this, depending on the system, only a certain object of interest among a plurality of objects is rendered by high-quality clipping or shading, and for other objects, a rectangular parallelepiped indicating a wire frame or an occupied area is rendered. There was a thing to simplify and render by such as. However, at this time, even if the viewpoint or the object moves, the objects other than the target object are always displayed in a simplified form regardless of the position or direction, so that the target object and the objects existing around it are changed in the scene change. There was a problem that it was not possible to clearly understand the situation.

According to the first aspect of the present invention, clipping or shading processing for each object is performed depending on whether or not each object is included in the angle range based on the direction of the object of interest from the viewpoint in the three-dimensional virtual space. By dynamically switching the content, the rendering method of each object can be appropriately and efficiently switched and set, and the computational load on the computer can be effectively reduced.

According to the second aspect of the present invention, the processing contents of clipping and shading for each object are dynamically changed depending on whether or not each object is included in the space area based on the object of interest in the three-dimensional virtual space. By switching, the rendering method of each object can be appropriately and efficiently switched and set, and the calculation load of the computer can be effectively reduced.

[0045]

In order to achieve the above object, the invention according to claim 1 of the present invention includes whether each object is included in an angle range based on a target object direction from a viewpoint in a three-dimensional virtual space. By having a means to dynamically switch the processing contents of clipping and shading for each object depending on whether or not it becomes possible to switch the rendering method of each object appropriately and efficiently,
Furthermore, it operates to reduce the computational load on the computer.

In order to achieve the above object, the invention according to claim 2 of the present invention is characterized by clipping or clipping each object depending on whether or not each object is included in the spatial region based on the object of interest in the three-dimensional virtual space. By providing a means for dynamically switching the processing content of shading, it becomes possible to appropriately and efficiently switch and set the rendering method of each object, and further operate to reduce the computational load on the computer.

Further, in more detail, the present invention can solve the above problems by the following constitution.

(1) An image generating apparatus for generating a two-dimensional image representing a scene (scene) of the space based on the object set in the three-dimensional virtual space, the light source, and the viewpoint, and creating model data of the object. Modeling means for converting the model data into a coordinate system of a viewpoint, a projection converting means for projecting model data based on the viewpoint coordinate system onto a projection surface, and an angle range based on a target object direction from the viewpoint. And a rendering means for switching the rendering method of each object.

(2) An image generating apparatus for generating a two-dimensional image representing a scene (scene) of the space based on the object, the light source, and the viewpoint set in the three-dimensional virtual space, and creating model data of the object. Modeling means, coordinate conversion means for converting the model data into a viewpoint coordinate system, projection conversion means for projecting model data according to the viewpoint coordinate system onto a projection surface, and rendering of each object based on a spatial region based on the position of the object of interest. Rendering means to switch methods,
An image generating apparatus comprising:

(3) Projection conversion means for mapping the model data of the object on the projection surface, clipping means for eliminating the surface of the object unnecessary for drawing based on the view volume (image forming area) based on the viewpoint, and the object. Comprising hidden surface elimination means for comparing the depths of the constituent surfaces to display the most front surface, shading means for shading the object surface, and texture mapping means for mapping a two-dimensional image on the object surface. The image generation apparatus described in (1) above.

(4) The image generating apparatus according to (1), wherein the content of the clipping means for each object is changed based on an angle range based on the direction of the object of interest from the viewpoint.

(5) The image generating apparatus according to (1), wherein the content of the shading means of each object is changed based on the angle range with respect to the direction of the object of interest from the viewpoint.

(6) The image generating apparatus according to (1), characterized in that the contents of the texture mapping means of each object are changed based on an angle range based on the direction of the object of interest from the viewpoint.

(7) Projection conversion means for mapping the model data of the object on the projection plane, clipping means for eliminating the surface of the object unnecessary for drawing based on the view volume (image forming area) based on the viewpoint, and the object Comprising a hidden surface removing means for comparing the depths of the surfaces forming the surface and displaying the surface in the foreground, a shading means for shading the object surface, and a texture mapping means for mapping a two-dimensional image on the object surface. The image generation apparatus according to (2) above.

(8) The image generating apparatus according to (2), wherein the contents of the clipping means for each object are changed based on the spatial area based on the position of the object of interest.

(9) The image generating apparatus according to (2), wherein the contents of the shading means of each object are changed based on the spatial area based on the position of the object of interest.

(10) The image generating apparatus according to (2), wherein the content of the texture mapping means of each object is changed based on the spatial area based on the position of the object of interest.

[0058]

BEST MODE FOR CARRYING OUT THE INVENTION (Embodiment 1) This embodiment relates to an image generating apparatus according to claim 1.

FIG. 1 is a diagram showing a system configuration in this embodiment.

In the figure, reference numeral 102 is a data input device for inputting initial values such as model data of an object and viewpoint conditions. For example, a keyboard is used.

Reference numeral 103 is a command input device for instructing the start / end of image generation processing, and for example, a mouse is used.

Reference numeral 104 is a file device for storing data such as a model, viewpoint, and illumination, and a nonvolatile memory is used.

Reference numeral 105 is an arithmetic unit for performing projection conversion using each value of the model, viewpoint, and illumination, each rendering process of clipping, hidden surface removal, shading, and updating of various variables.

Reference numeral 106 is a memory device for storing various data such as model data of the object, viewpoint conditions, illumination conditions, etc. during the activation of the device.

Reference numeral 107 denotes a data display device 107 for displaying an instruction to the user from the control device 101, input data and the like, and for example, a TFT liquid crystal monitor is used.

Reference numeral 108 is a video memory device for storing the pixel values of the image after rendering.

An image display device 109 for displaying an image on the video memory device 108 is, for example, C.
RT (Cathode Ray Tube)
A monitor is used.

A control device 101 controls the data input device 102, the command input device 103, the file device 104, the arithmetic device 105, the memory device 106, the data display device 107, and the image display device 108.

After the image generation system is activated, first, initial data of a plurality of models (vertex data and surface loop data), viewpoints, and light sources stored in the file device 104 are loaded into the memory device 106. Next, the user of the system inputs the identification number of the target object from the keyboard which is the data input device 102. When the input is finished, the system starts a loop of image generation processing.

At the beginning of the image generation processing loop, the system
Based on the reference position of each object at that time and the angle of view for the object of interest (meaning the angle range based on the direction of the object of interest from the viewpoint position), each object is selected as a normal display object or a simple display object. Classify. The system then uses this classification result to render each object. That is, for objects classified as normal display objects, rendering is performed with normal clipping and shading, and for objects classified as simple display objects, clipping and shading are simplified and rendering is performed. In the process, rasterization (substitution of pixel values) of each pixel forming a display image is performed in the video memory device 108. The display image stored in the video memory device 108 is the image display device 1.
It is output from 09 and is displayed. Note that the objects classified as non-display objects are not rendered.

Next, the system updates the position of each object and viewpoint. That is, it moves by a minute distance Δd in the direction indicated by the direction vector of each object or viewpoint. However, when the position of the object or the viewpoint is outside the movable area, the position and direction of the object or the viewpoint are appropriately newly determined so as not to exceed the movable space (in the present embodiment, the movable area is set as the movable area). , Consider a spherical region of radius R centered at the origin).

At the end of the image generation processing loop, the system determines whether or not there is a command from the user to end the image generation processing, and if so, terminates the image generation processing, otherwise returns to the beginning of the loop.

Regarding the image generation processing executed by the image generation system configured as described above, the processing contents will be described according to the flow of data.

FIG. 15 is a flow chart for explaining the processing of the image generating apparatus in this embodiment. The outline of the control executed by the image generating apparatus configured as shown in the figure will be described below according to the flow of data.

Step S1 After starting the image generation system, first, the object data is initialized.

First, the control device 101 is the file device 1
The model data (for example, the one in FIG. 4) configured by the vertex data and the surface loop data stored in 04 is stored in the memory device 106 as a data list for objects (data for 10 objects 0 to 9 shown in FIG. 12). It is loaded to each model data area of (). Also, a texture image (for example, FIG.
No. 1) is loaded from the file device 104 and stored in each texture image data area in the object data list. It is assumed that the texture image for each object is created in advance by a paint system or the like and is stored in the file device 104.

As the initial position of each object, the position in the movable area in the three-dimensional virtual space is randomly calculated and set in each position data area in the object data list. However, in the present embodiment, a spherical region having a radius of 25.0 (above R) centered on the origin of the three-dimensional space is considered as the movable region, but the present invention is not limited to this.

A three-dimensional normalized vector (normalized vector means a vector having a length of 1) is randomly calculated as the initial direction of each object and set in each direction data area of the object data list. To be done.

Step S2 Here, the viewpoint condition is initialized.

That is, the control device 101 uses the viewpoint data file in the file device 104 to generate a data set shown in FIG. 13, for example, viewpoint position (cx, cy, cz): (10.0, 5.0, 8.0) viewpoint direction (cdx , Cdy, cdz): (1.0, 1.0, 0.0) Angle (α, β): (4π / 16, 3π / 16) Distance to each clipping plane of near and fur (distN, distF): (1.0, 100.0) The viewing angle for the object of interest (α ′, β ′): (4π / 32, 3π / 32) is loaded and set in the memory device 106.

Here, α, β, α'and β'refer to the following angles in FIG.

Α = ∠KPM β = ∠LPN α '= ∠FPH β '= ∠EPG Step S3 Here, the light source conditions are initialized.

That is, the control device 101 loads the data set shown in FIG. 14, for example, the following values from the light source data file in the file device 104, and sets the data in the memory device 106.

Light source direction (lx, ly, lz): (-1.0, -1.0, -1.0) Light source intensity I: 10.0 However, in this embodiment, a parallel light source (light source set to infinity) is considered. There is no light source position.

Step S4 Next, the control device 101 displays a message prompting the input of the identification number of the object of interest on the liquid crystal display which is the data display device 107. Please input the identification number of the object of interest >>>>. In this state, the user inputs the identification number of the target object from the keyboard which is the data input device 102. The input identification number is substituted into the variable targetID in the memory device 106.

When the user finishes inputting the identification number of the object of interest, the control device 101 starts the loop of the image generation process from the next step.

Step S5 At the beginning of the image generation processing loop, the rendering method of each object is set.

FIG. 16 shows a method of setting the rendering method for each object.

The arithmetic unit 105 is instructed by the control unit 101.
Is the position of the object of interest (identification number tar
0 for the rendering mode variable rmodei of an object included in the angular range of left and right α ′ and upper and lower β ′ around the direction toward the position of the object having getID).
(Meaning of normal rendering) is substituted, and 1 (meaning of simple rendering) is substituted for the same variable for an object not included.

Step S6 Here, the rendering mode variable rmodei of the object
Is used to render each object.

That is, in accordance with the instruction of the control device 101, the arithmetic unit 105 switches the rendering method as follows according to the value of the rendering mode variable rmodei of each object in the memory device 105, and the arithmetic unit 105 displays each on the video memory device 108. Expands (rasterizes) the polygons that make up an object.

If the object's rmodei is 0, the object is rendered (clipping, shading,
Texture mapping is performed by the following normal method.

Clipping: The polygons forming the object are clipped by the usual method.

Shading: Smooth shading is performed.

Texture mapping: execute.

If the object's rmodei is 1, the object is rendered (clipping, shading,
Texture mapping) is performed by the following simple method.

Clipping: When each object position (object reference position) is included in the view volume, each polygon forming the object is clipped by a normal method. If not included, all polygons that form the object are excluded from the rendering target. As a result, for an object for which simple rendering is specified, when the object position (reference position of the object) is outside the view volume, the object does not draw even if a part of the object intersects the view volume. .

Shading: Flat shading is performed.

Texture mapping: Do not execute.

However, in this rendering process,
As a matter of course, drawing is not executed for an object (306 non-rendering object in FIG. 16) that does not intersect the view volume at all.

At this step, the video memory device 108
The two-dimensional image created above is passed to the image display device 109 and displayed.

As an example of the generated image, a two-dimensional image obtained by rendering the scene in the three-dimensional space shown in FIG. 16 is shown in FIG.

In this figure, four objects arranged around the object of interest are subjected to smooth shading and texture mapping, and other objects are subjected only to flat shading.

Step S7 Here, the position of each object and the viewpoint is updated.

That is, according to the instruction from the control device 101, the arithmetic device 105 moves the current position of each object in the direction indicated by the direction vector of the object by a minute distance Δd. Similarly, the current viewpoint position is moved by a minute distance Δf in the direction indicated by the viewpoint direction vector. However, if the position of the object (or viewpoint) is outside the spherical movable area, the new position is set to the nearest coordinate on the area boundary. At this time, the new direction of the object (or viewpoint) is set appropriately, that is, at random so as to face the inside of the spherical region.

Step S8 At the end of the image generation processing loop, the control device 101 determines whether or not the user has instructed the command input device 103 to end the image generation processing, and if there is, terminates the image generation processing. Return to S5.

(Embodiment 2) This embodiment relates to the image generating apparatus according to the second aspect.

The system configuration of this embodiment is shown in FIG. 1 as in the case of the first embodiment.

The flow of the image generation processing in this embodiment is shown in FIG. 15 as in the first embodiment, but objects are classified based on the sphere area centered on the position of the object of interest in the three-dimensional virtual space. Is different. Along with this, the processing contents of steps S2 and S5 in the processing flow are changed to the following steps S2 ′ and S5 ′, respectively.

Step S2 ' Here, the viewpoint condition is initialized.

That is, the control device 101 uses the viewpoint data file in the file device 104 to generate a data set shown in FIG. 18, for example, viewpoint position (cx, cy, cz): (10.0, 5.0, 8.0) viewpoint direction (cdx , Cdy, cdz): (1.0, 1.0, 0.0) Viewing angle (α, β): (4π / 16, 3π / 16) Distance to each clipping plane of near and fur (distN, distF): (1.0, 100.0) The target object spherical region radius targetR: 5.0 is loaded, and setting is performed for each data region of the viewpoint data set in the memory device 106.

Here, α and β are the following angles in FIG.

Α = ∠KPM β = ∠LPN Step S5 ′ At the beginning of the image generation processing loop, the rendering method of each object is set.

FIG. 19 shows a method of setting the rendering method for each object.

That is, the arithmetic unit 105 is instructed by the control unit 101 as shown in FIG.
Radius target centered on the position of (object with ID)
For an object included in the R boundary area, 0 for the rendering mode variable rmodei of this object (normal display)
Substituting, and for objects not included, 1 (meaning simple display) is assigned to the same variable.

FIG. 20 shows an example of a two-dimensional image generated in the process of image generation processing in this embodiment. The image in this figure is a rendering of the scene in the three-dimensional space shown in FIG. In FIG. 20, four objects arranged around the object of interest are subjected to smooth shading and texture mapping, and other objects are subjected only to flat shading. In addition, 3 in FIG.
The arrangement of objects and viewpoints in the dimensional space is exactly the same as in the case of FIG. 16, but the rendering content of the generated two-dimensional image is different.

That is, compared to the two-dimensional image of FIG. 17, the occupied areas of each object in the image are the same in FIG. 20, but the target objects of normal rendering and simple rendering are different.

[0118]

As described above, claim 1 of the present application
According to the invention related to the above, it is generated by including means for dynamically switching the rendering processing contents of clipping and shading for each object based on the angle range based on the direction of the object of interest from the viewpoint in the three-dimensional virtual space. On the three-dimensional image, the object of interest and the objects placed around it can be displayed in high quality, and other objects can be displayed easily, and the operation load can be reduced efficiently by the computer. became.

Furthermore, there is an additional effect that the object of interest and the objects around it can be highlighted.

According to the invention of claim 2 of the present application,
By having means for dynamically switching the rendering processing contents of clipping and shading for each object based on the spatial region based on the object of interest in the three-dimensional virtual space, the object of interest and its surroundings can be displayed on the generated two-dimensional image. The objects to be placed can be displayed in high quality, and other objects can be displayed easily, and the operation to reduce the calculation load of the computer efficiently has come to be performed.

Furthermore, there is an additional effect that the object of interest and objects around it can be emphasized and displayed.

[Brief description of drawings]

FIG. 1 is a diagram showing a system configuration of Examples 1 and 2.

FIG. 2 is a diagram showing the flow of general two-dimensional image generation in three-dimensional CG.

FIG. 3 is a diagram showing a three-dimensional object in a modeling coordinate system.

FIG. 4 is a diagram showing an example of model data.

FIG. 5 is a diagram showing four coordinate systems for projection conversion.

FIG. 6 is a diagram showing projection conversion.

FIG. 7 is a diagram showing clipping processing in rendering, in which (a) is a perspective view of the clipping processing state;
(B) is a front view of the projection surface of the clipping process of (a)

FIG. 8 is a diagram showing a shading model.

FIG. 9 is a diagram showing how to obtain the directions of vertices.

FIG. 10 is a diagram showing a method of smooth shading.

FIG. 11 is a diagram showing a texture mapping process by perspective coordination.

FIG. 12 is a diagram showing an object data list.

FIG. 13 is a diagram showing a viewpoint data set according to the first embodiment of the present invention.

FIG. 14 is a diagram showing a light source data set.

FIG. 15 is a diagram showing a flowchart of Examples 1 and 2.

FIG. 16 is a diagram showing a method of determining a rendering method for an object according to the first embodiment of the present invention.

FIG. 17 is a diagram showing an example of a generated image according to the first embodiment of the present invention.

FIG. 18 is a diagram showing a viewpoint data set according to the second embodiment of the present invention.

FIG. 19 is a diagram showing a method of determining a rendering method for an object according to the second embodiment of the present invention.

FIG. 20 is a diagram showing an example of a generated image according to the second embodiment of the present invention.

[Explanation of symbols]

101 control device 102 data input device 103 command input device 104 file device 105 arithmetic device 106 memory device 107 data display device 108 video memory device 109 image display device 301 viewpoint 302 target object 303 based on the direction of the target object viewed from the viewpoint Angle range 304 Normal rendering target object 305 Simple rendering target object 306 Non-rendering object 307 Projection plane (near clipping plane) 308 Axis of view 309 Direction of target object from viewpoint 401 Target object 402 Normal rendering object 403 Simple rendering object 501 View point 502 Object of interest 503 Spherical region 504 centered on the object of interest 504 Normal rendering target object 505 Simple rendering target object 506 Non-rendering object 507 Projection plane (near clipping plane) 50 Visual axis 601 focuses the object 602 typically render object 603 simple rendering object

   ─────────────────────────────────────────────────── ─── Continued front page    F-term (reference) 5B050 AA03 BA09 BA11 BA18 EA07                       EA28 EA29 EA30 FA02                 5B080 BA01 BA03 BA05 BA07 GA11                       GA22

Claims (10)

[Claims] 1. An image generation apparatus for generating a two-dimensional image representing a scene (scene) in the space based on the object set in a three-dimensional virtual space, a light source, and a viewpoint, and creating model data of the object. Modeling means, coordinate transformation means for transforming the model data into a coordinate system of a viewpoint, projection transformation means for projecting model data according to the viewpoint coordinate system onto a projection surface, based on an angle range based on a target object direction from the viewpoint And a rendering unit that switches a rendering method of each object. 2. An object, a light source, set in a three-dimensional virtual space,
And an image generation device for generating a two-dimensional image representing a scene (scene) in the space based on a viewpoint, modeling means for generating model data of an object, and coordinate conversion for converting the model data into a coordinate system of the viewpoint. Image generation, which comprises: means, projection conversion means for projecting model data based on the viewpoint coordinate system onto a projection surface, and rendering means for switching a rendering method of each object based on a spatial area based on the position of the object of interest. apparatus. 3. Projection conversion means for mapping model data of an object onto a projection surface, clipping means for eliminating a surface of the object unnecessary for drawing based on a view volume (drawing area) based on the viewpoint, and the object are configured. It is characterized by comprising hidden surface elimination means for comparing the depths of the surfaces to be displayed and displaying the frontmost surface, shading means for shading the object surface, and texture mapping means for mapping a two-dimensional image on the object surface. The image generation apparatus according to claim 1. 4. The image generating apparatus according to claim 1, wherein the content of the clipping means for each object is changed based on an angle range based on the direction of the object of interest from the viewpoint. 5. The image generating apparatus according to claim 1, wherein the content of the shading means of each object is changed based on an angle range based on the direction of the object of interest from the viewpoint. 6. The image generating apparatus according to claim 1, wherein the content of the texture mapping means of each object is changed based on an angle range based on the direction of the object of interest from the viewpoint. 7. A projection conversion means for mapping model data of an object onto the projection surface, a clipping means for eliminating a surface of the object unnecessary for drawing based on a view volume (drawing area) based on the viewpoint, and an object. Comprising hidden surface elimination means for comparing the depths of the constituent surfaces to display the most front surface, shading means for shading the object surface, and texture mapping means for mapping a two-dimensional image on the object surface. The image generation apparatus according to claim 2, characterized in that 8. The image generating apparatus according to claim 2, wherein the content of the clipping means of each object is changed based on the spatial area based on the position of the object of interest. 9. The image generating apparatus according to claim 2, wherein the content of the shading means of each object is changed based on the spatial area based on the position of the object of interest. 10. The image generating apparatus according to claim 2, wherein the content of the texture mapping means of each object is changed based on the spatial area based on the position of the object of interest.