JPH09153142A - Device and method for rendering - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently execute rendering processing by converting each intensity of RGB given as chrominance information of the origin of a polygon to luminance information and color difference information, thereby reducing a processing amount necessary at the time of executing rendering processing for displaying the object of graphics. SOLUTION: At first, after acquiring the RGB value of the vertex of the polygon to be the object of plotting, the RGB value of the vertex is converted into a YUV value. Next based on the YUV value of the vertex of the polygon, the YUV values of all the pixcels constituting the polygon is calculated. At last, the YUV value is stored in a frame buffer 15. In this case, processing converting the RGB value into the YUV value should be executed by only the number of the vertexes of the polygon. As the number of the vertexes of the polygon is smaller than the number of the pixcels constituting the polygon and the difference becomes large along with the increase of the size of the polygon, the processing converting the RGB value into the YUV value is drastically reduced.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a high-speed rendering process in the field of computer graphics systems, and more particularly to a rendering technique suitable for performing various graphics operations.

[0002]

2. Description of the Related Art In a device that displays color graphics, such as a workstation or a PC, in order to display a three-dimensional shape, the shape is divided into surfaces having triangular or quadrangular shapes called polygons, and then , R, G, and B intensities of all pixels forming the polygon are calculated, and finally, the calculated RGB intensities are stored in a device for storing an image, that is, a frame buffer. Further, in order to generate an effect of making a drawn object transparent or semi-transparent, an A (alpha) value indicating transparency may be calculated in the same manner as the R, G, B values and stored in the frame buffer. .

Here, assuming that each value of R, G, B, and A indicating the color of a pixel forming an image is represented by 8 bits, and the display screen is composed of 640 × 480 pixels,
The capacity of the frame buffer is 640 × 480 × 8 × 4 bits = about 1 gigabyte, which is a huge amount. As the capacity of the frame buffer increases, the cost, base area, and power consumption of the memory for the frame buffer increase, and in order to improve drawing performance, it is necessary to speed up writing to the frame buffer. Is coming.

To solve this problem, Japanese Patent Laid-Open No. 61-
According to the one described in Japanese Patent No. 233781, after calculating the respective intensities of R, G, and B of all the pixels constituting the polygon, the RGB value of each pixel is converted into the luminance information and the color difference information to obtain the luminance information. And the color difference information are stored in the frame buffer. At that time, by storing the color difference information of one pixel in the frame buffer as a representative of the color difference information of a plurality of pixels, the amount of color difference information stored in the frame buffer is reduced, and the capacity of the entire frame buffer is suppressed. ing. Therefore, it is possible to prevent an increase in the cost of the memory, the substrate area, and the power consumption, and to mitigate the deterioration of the drawing performance due to the writing process to the frame buffer.

The technique for reducing the color difference information is that "the resolution of the human visual sense when the color is changed while the luminance is constant is inferior to that when the color is constant and the luminance is changed. Is based on the fact that
System color TV and M, one of the moving image compression technologies
It is also used in PEG.

[0006]

However, even if the capacity of the frame buffer is reduced and the deterioration of the drawing performance due to the writing process to the frame buffer is suppressed, the intensities of RGBA of the pixels forming the polygon are calculated. Since the amount of calculation of rendering processing such as shading, texture mapping, and fog performed on the computer is enormous, it is unavoidable that the rendering performance will be degraded unless these rendering processings are executed efficiently.

The rendering process will be briefly described. Shading is a process for providing a smooth shading, and texture mapping prepares a two-dimensional pattern (texture) and attaches it to a polygon to improve the image reality. Is. In addition, fog can represent a state where the entire screen is blurred, and fog, haze, and air pollution can be reproduced. For more information about the rendering process, see "" Open PenGLogramming Guide (Japanese version) ".
(Addison Wesley, published by Seiunsha Co., Ltd.) "and the like.

Whether or not these rendering processes are executed efficiently has a great influence on the drawing performance.
When displaying a complicated shape including a curved surface, the shape is divided into thousands to hundreds of thousands of polygons and RGB intensities are calculated, so that the influence of rendering processing increases. . In addition, a VR (Virtual R) that enables a virtual object existing in a virtual space defined in a computer to be operated as if it were real
Even when using the (eality) technology, drawing performance close to real time is required, and therefore, it is necessary to provide a means for calculating the color of the pixel inside the polygon forming the three-dimensional shape at high speed while suppressing the deterioration of image quality. .

The problem to be solved by the present invention is to reduce the processing amount required for executing the rendering process for displaying the object shape of the graphics in the case of displaying color graphics, thereby performing the rendering process. Is to execute efficiently.

[0010]

A rendering device of the present invention which solves this problem is a graphics rendering device for calculating color information of internal pixels of a polygon called polygon, which constitutes a shape to be drawn. Then, the brightness information and the color difference information of the pixel inside the polygon are calculated from the conversion means for converting the R, G, and B intensities of the pixel into the brightness information and the color difference information, and the brightness information and the color difference information of each vertex of the polygon. And a storage unit that stores the brightness information and the color difference information. The conversion unit converts each intensity of RGB given as the color information of the vertices of the polygon into the brightness information and the color difference information. Based on the luminance information and the color difference information acquired from the converting means as the color information of the vertex of the polygon, the luminance information and the color difference information of the pixels inside the polygon are calculated, and the luminance information and the color difference information are calculated. And performing processing of storing in the storage means.

Further, the calculating means calculates the color difference information of the pixel inside the polygon by using the color difference information of one pixel for the plurality of pixels and the transparency information as a representative of the plurality of color difference information, and stores the color difference information. The means is one for a plurality of pixels
A rendering device that stores the color difference information of a pixel as a representative of the plurality of color difference information is preferable.

Further, the computing means, based on the transparency information of each vertex of the polygon, for the plurality of pixels, the transparency information of one pixel is represented as a representative of the plurality of transparency information.
A rendering device that calculates transparency information of pixels forming a polygon and stores the transparency information in the storage means as a representative of the plurality of transparency information, or a computing means, depth information of each vertex of the polygon. Based on the above, the depth information of pixels forming a polygon is calculated by using the depth information of one pixel as a representative of the plurality of depth information for a plurality of pixels, and the depth information is calculated as the depth information of the plurality of depths. A rendering device that performs a process of storing in the storage unit as a representative of information, or a calculation unit, based on the fog information of each vertex of the polygon, the fog information of one pixel for a plurality of pixels. The fog information of pixels forming a polygon is calculated as a representative of the individual fog information, and the depth information is used as a representative of the plurality of fog information. Also preferred rendering device that performs a process of storing the 憶 means.

First, the converting means is given the respective intensities of R, G, and B of the vertices of the polygons forming the shape to be displayed.

Then, the converting means converts the given RGB values into luminance information and color difference information, and outputs the luminance information and color difference information to the calculating means.

The calculating means calculates the luminance information and the color difference information of the pixels forming the polygon. That is, based on the luminance information and the color difference information of each vertex of the polygon received from the conversion means, shading,
Rendering processing such as texture mapping and fog is executed to calculate the luminance information and color difference information of pixels, and the luminance information and color difference information are stored in the storage means.

Further, in order to draw a transparent or semi-transparent polygon, transparency information is calculated by a rendering process like the brightness information and color difference information, and the transparency information is stored in a storage means, or three-dimensional graphics. In some cases, the depth information is calculated and the depth information is stored in the storage means in order to execute the hidden surface processing in. In addition, fog information may be calculated in order to execute fog processing in the rendering processing.

Here, the intensity of the RGB value is converted into luminance information and color difference information before executing the rendering process, and 1
By executing the rendering process by using the color difference information of the pixel as a representative of the plurality of color difference information, the number of color elements to be calculated per pixel is reduced, and the efficiency of the entire rendering process can be improved. Further, as in the case of the color difference information, if the rendering processing is executed by using the transparency information, depth information, and fog information of one pixel as a representative of the transparency information, depth information, and fog information of a plurality of pixels, the effect is obtained. Higher.

Further, the brightness information of each pixel calculated by the rendering process, the color difference information and the transparency information that are representative of the plurality of pixels are stored in the frame buffer, and the depth information that is the representative of the plurality of pixels is stored in the Z buffer. By storing, the capacity of the frame buffer and the Z buffer can be reduced.

Furthermore, by performing the rendering process using the YUV value, the MPE has an additional effect.
It becomes easy to execute texture mapping in which G data is attached to polygons. In other words, MPEG data is YUV
Since the video data including the values is compressed, the moving image can be attached to the polygon without requiring a special process of converting the YUV value into the RGB value. Therefore, MP
Since the live-action image is mainly compressed in the EG data,
Fusion of CG and live action can be realized more easily.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of a graphics rendering system which is an embodiment of the present invention.

The rendering system shown in FIG. 1 includes graphic information such as vertex coordinates of a polygon, R of the vertex of the polygon,
A computer 11 which owns graphic data 12 consisting of color information indicating the intensities of G and B, a rendering device 20 which calculates the color of the pixels forming a polygon, a decoder 16 which converts YUV values into RGB values, and a digital device. It has a D / A converter 17 for converting a signal into an analog signal and a CRT 18 for displaying a rendering result.

Further, the rendering device 20 sets the RGB value to 1
An encoder 13 that converts one luminance value Y and two color difference values U and V, a graphics processor 14 that calculates the YUV values of all the pixels that make up the polygon from the YUV values of the vertices of the polygon, and the YUV values are stored. The frame buffer 15 is provided.

Next, the operation of the rendering system having such a system configuration will be described.

First, the computer 11 encodes the RGB values of the given graphic data 12 in order to display the graphic.
The graphic information including the vertex coordinates is output to the graphics processor 14 at 13. The encoder 13 that has acquired the RGB value converts the RGB value into one luminance value Y and two color difference values U and V, and outputs the result to the graphics processor 14. Conversion from RGB values to YUV values is generally done by "CCIR
Recommended 601 ”, and R, G, B, Y,
When each of U and V is expressed in 256 steps of 8 bits, the conversion formula from RGB to YUV is as follows.

[0026]

[Formula 1] Y = 0.299R + 0.587G + 0.114B U = −0.1687R−0.3313G + 0.5B + 128 V = 0.5R−0.4187G−0.0813G + 128 (Formula 1) Actually, the processing speed is considered. In many cases, the RGB values are converted into YUV values by converting each coefficient into an integer type and then performing addition / subtraction of integers without using the real number type operation.

Next, the graphics processor 14, which has received the YUV value and the graphic information, executes a rendering process based on the information, and calculates the YUV value and the coordinate value of all the pixels forming the polygon. Rendering processing includes shading, texture mapping, and alpha blending. At that time, since the color components of the pixels are represented by YUV values instead of RGB values, the rendering method applied to the pixels is also executed for each of the Y, U, and V components. For example, in the shading process, y
While keeping the coordinates constant and increasing the x coordinate by 1 to determine the color of the pixel, conventionally, for each component of R, G, B,

[0028]

## EQU00002 ## R = R + dR / dx G = G + dG / dx B = B + dB / dx (Equation 2) Instead of executing

[0029]

## EQU00003 ## Y = Y + dY / dx U = U + dU / dx V = V + dV / dx (Equation 3) is executed. Then, the YUV value calculated by the graphics processor 14 is stored in this position in the frame buffer 15, which is determined based on the pixel coordinates (x, y) derived from the graphic information.

The decoder 16 synchronizes with the scanning of the CRT 18,
CRT the YUV value stored in the frame buffer 15
Restore RGB values for display at 18. The conversion formula from the YUV value to the RGB value is derived from Expression 1 as Expression 4.

[0031]

## EQU00004 ## R = Y + 1.402 (V-128) G = Y-0.34414 (U-128) -0.71414 (V-128) B = Y + 1.772 (U-128) (Equation 4) Also in this case, as in the case of the equation 1, considering the processing speed,
In many cases, each coefficient is converted into an integer type and an integer addition / subtraction multiplication is executed without using a real number type operation.

The RGB value calculated from equation 4 is D
Converted to analog information by the A / A converter 17 and C
It is sent to RT18. The CRT 18 displays the analog information received from the D / A converter 17 on the display.

In the rendering system to be described, the graphics processor 14 calculates the brightness information and the color difference information of the pixels forming the polygon, but the computer 11 does not cause any problem.

FIG. 2 is a block diagram of a conventional graphics rendering system.

The components of the rendering system shown in FIG. 2 are the same as those of the rendering system shown in FIG. 1, except that the positions of the graphics processor 14 and the encode 13 are different.

Next, the operation of the rendering system having such a system configuration will be described.

First, the computer 11 outputs the RGB values of the given graphic data 12 and the graphic information to the graphics processor 13. Graphics processor 14a
Rendering processing is executed based on the received RGB values of the vertices of the polygon and the graphic information, and the RGB values and the coordinate values of all the pixels forming the polygon are calculated.

Next, the encoder 13 uses Equation 1 to convert the RGB values of the pixels forming the polygon into YUV values,
The result is written in the frame buffer 15. Decoder 16
The RGB values calculated by using the equation 4 are converted into analog information by the D / A converter 17 and sent to the CRT 18. The CRT 18 displays the analog information received from the D / A converter 17 on the display.

Here, in order to compare the conventional method with the present invention, a flowchart of the rendering process by the conventional method is shown in FIG. 3, and the rendering executed by the encoder and the graphics processor shown in FIG. 1 is performed. FIG. 4 shows a flowchart of the program when the processing is executed by software.

In the conventional method, as shown in FIG.
First, after obtaining the RGB values of the vertices of the polygon to be drawn (step 30), rendering processing such as shading is executed to obtain the RGB values of the pixels inside the polygon.
Calculate the value (step 31). Next, the calculated RGB values of all the pixels inside the polygon are converted into YUV values (step 32). Finally YUV derived by conversion
The value is stored in the frame buffer (step 33). In this method, it is necessary to perform conversion processing from RGB values to YUV values for the number of pixels that form a polygon.

According to the present invention, as shown in FIG.
First, after obtaining the RGB values of the vertices of the polygon to be drawn (step 40), the RGB values of the vertices are converted into YUV values (step 41). Next, the Y of the vertex of the polygon
YU of all the pixels that make up the polygon based on the UV value
The V value is calculated (step 42). Finally, the YUV value is stored in the frame buffer (step 43). In this case, the process of converting the RGB value into the YUV value may be executed for the number of vertices of the polygon. The number of vertices of a polygon is less than the number of pixels that make up the polygon, and the larger the size of the polygon, the more significant the difference becomes.
The conversion processing from RGB values to YUV values can be greatly reduced.

Further, in the present invention, when the texture mapping process, which is one of the rendering processes, is executed,
If image data to be attached to a polygon, which is called a texture, is given as a YUV value, it is sufficient to extract the YUV value of the texture corresponding to the pixel of the polygon.
The conversion process to the GB value can be deleted. Therefore,
Moving image data compressed in the YUV value format, for example MPEG which is becoming a standard moving image compression format.
If the data is used as moving image data, the moving image can be attached to the polygon simply by expanding the MPEG data. That is, if the moving image data created by capturing the moving image data is used, it is easy to combine computer graphics (CG) and the actual image.

Here, in the apparatus shown in FIG. 1, the screen of the CRT is arranged in the vertical and horizontal directions for each predetermined number of pixels, for example, in the vertical direction.
It is divided into 2 horizontal pixels, and the U and V values are 2 × for each divided rectangle.
Consider a case where the rendering process is executed as a representative value of 2 pixels and the result is stored in the frame buffer. That is, with respect to the display image shown in FIG. 4, the Y value is the pixel p0 (100) as shown in FIG.
Has a luminance value Y0 (200), and the pixel p1 (101) has a luminance value Y1.
(201) corresponds to the brightness value Y2 (202) to the pixel p2 (102) and corresponds to the brightness value Y3 (203) to the pixel p3 (103), and one brightness value is assigned to each pixel. On the other hand, in the case of U value, for the display image shown in FIG.
Corresponds as shown in FIG. That is, the pixel p0 (10
0), p1 (101), p2 (102), p3 (103) have color difference value U0
(300) is the pixel p4 (104), p5 (105), p6 (106), p7
The color difference value U4 (304) corresponds to (107), one U for every 4 pixels.
Value is assigned. Similarly for V value, 1 for 4 pixels
V values are assigned. When each of the U and V values is stored in the frame buffer, writing can be easily executed by setting the least significant bit of the pixel coordinate values X and Y to 0 and determining the storage destination address.

As a result, the spatial resolution of U and V is reduced to 1/4.
Since the capacity of the frame buffer is reduced to 1/2 of that of the conventional one, it is possible to suppress an increase in memory cost, base area, and power consumption. Further, since the amount of data written in the frame buffer is halved, the drawing performance is also improved.

Here, taking as an example the case where the shading process in graphics is executed, a comparison is made between the case where the spatial resolution of the color difference values is not reduced at all and the case where the spatial resolution of the color difference values is reduced to 1/4.

Although there are many algorithms for the shading process, first, a representative algorithm will be briefly described here.

As shown in FIG. 8, the triangle shading process is composed of an edge process 25 and a span process 26. In the edge process 25, coordinates (x, y, z) on AB, which is one side of the triangle, are used. And color (R, G, B) or (Y,
U, V) are sequentially obtained, and at the same time, the span processing 26
Coordinates and colors are calculated sequentially for each line.

Next, in order to make a comparison between the case where the spatial resolution of the color difference value is not reduced and the case where it is dropped, one Y, U, and V are assigned to one pixel, respectively. FIG. 9 shows a flowchart of the shading process, and FIG. 1 shows the shading process in graphics when one is assigned to four pixels for U and V.
0 is shown.

First, referring to FIG. 9, Y, U, and
A shading process in graphics when one V value is assigned to each will be described.

First, the coordinates of the vertices of the polygon to be drawn and the YUV value are acquired (step 50). Next, it is judged whether or not the calculation of the YUV values of the pixels forming the polygon has been completed, and if completed, the process ends (step 5).
1). Otherwise, the x coordinate and Y during edge processing
The UV value is saved (step 52), each time the calculation of one line is completed, 1 is added to the y coordinate and the increment value is added to the x coordinate and the YUV value (step 56), and the next line is processed. Move. The above is the edge processing. If the calculation of one line is not completed (step 53), the YUV being calculated
Store the value in memory (step 54) and set the x coordinate to 1
The process of adding the increment value to the YUV value (step 55) is 1
Repeat until line calculation is completed. The above is the span processing.

Next, with reference to FIG. 10, a shading process in graphics when one U and V value is assigned to four pixels will be described.

Since the rendering process is executed simultaneously for four pixels, there are four variables (Y0, Y1, Y2, Y) for storing the brightness value.
3) Prepared. Further, since the 4 pixels are a rectangle of 2 pixels in the vertical direction and 2 pixels in the horizontal direction, the span processing is executed simultaneously for two lines.

First, the coordinates of the vertices of the polygon to be drawn and the YUV value are acquired (step 60). Next, it is judged whether or not all the YUV values of the pixels forming the polygon have been calculated, and if the calculation is completed, the process ends (step 61). Otherwise, x being calculated by edge processing
The coordinates and the YUV value are saved (step 62), and the y coordinate is set to 2 each time the calculation of the YUV values of the pixels for two lines is completed.
Is added, and the increment value is added to the x coordinate and the YUV value (step 66). If the calculation of the YUV value of the pixels for two lines has not been completed (step 63), the Y being calculated is being calculated.
Store UV value in memory (step 64) and set 2 to x coordinate
The process of adding the increment value to the YUV value (step 65) is repeated until the calculation for two lines is completed.

As a result, in the case of shading processing, 1
For each pixel of Y, U, V components
Since it was necessary to execute the number of times, by lowering the spatial resolution of U and V, in the shading process of 4 pixels, the first equation of Formula 3 was calculated four times for Y and several times for each color component of U and V. It suffices to perform the calculations of the second and third equations of 3 once. Therefore, the calculation amount of shading processing is about 1
It is reduced to / 2 and the drawing performance is improved.

FIG. 11 is a block diagram of another embodiment according to the present invention. The feature of the present embodiment is that the A (alpha) value of the parameter indicating the transparency is added to the color component of the pixel. .

The present configuration has a computer 11c which owns graphic data 12 including graphic information such as vertex coordinates of polygons and color information indicating the intensity of RGBA values, and a rendering device which calculates the YUVA values of the pixels forming the polygon. 20
And a decoder 16 for converting YUV values into RGB values, D /
It has an A converter 17 and a CRT 18.

Further, the rendering device 20 uses the encoder 13 for converting the RGB values into the YUV values and the YUVA value of each vertex of the polygon, and the YU of all the pixels forming the polygon.
Graphics processor 14c, YU for calculating VA value
A frame buffer 15c for storing the VA value is provided.

First, alpha blending, which is one of the rendering processes using the A value, for realizing the drawing of transparent and semitransparent polygons will be described.

Generally, the A (alpha) value used in alpha blending is used to mix the RGBA value stored in the frame buffer with the RGBA value during the rendering process. There are several ways to mix, but the most commonly performed formulas are shown below.

[0060]

## EQU00005 ## R = R * A + Rb * (1-A) G = G * A + Gb * (1-A) B = B * A + Bb * (1-A) A = A * A + Ab * (1-A) ... ( (5) Rb, Gb, Bb, Ab are RGBA values stored in the frame buffer, R, G, B, A are RGBA values during rendering processing, and A is a value in the range of 0 to 1. .

In this case, a certain object is drawn and then A
If a polygon having a value of 0.3 is drawn, 30% of the polygon color and 70% of the object color are mixed, and a semitransparent polygon can be displayed.

When alpha blending is applied not to R, G and B but to the converted Y, U and V components,

[0063]

(6) Y = Y * A + Yb * (1-A) U = U * A + Ub * (1-A) V = V * A + Vb * (1-A) A = A * A + Ab * (1-A) ( Equation 6) will be executed. Yb, Ub, Vb, and Ab are YUVA values stored in the frame buffer, Y, U,
V and A represent YUVA values during the rendering process. When executing the alpha blending process, it is necessary to store the A value together with the Y, U and V values in the frame buffer.

Next, the operation of the rendering system having the system configuration shown above will be described.

In order to draw a transparent or semi-transparent object, the computer 11c first sets the R of the given graphic data 12.
The GB value is output to the encoder 13, and the A value and graphic information including the vertex coordinates are output to the graphics processor 14c. RG
The encoder 13 that has acquired the B value converts the RGB value into the YUV value and outputs the result to the graphics processor 14c. The graphics processor 14c having received the YUVA value and the vertex coordinates calculates the YUVA value and the coordinate value of all the pixels forming the polygon based on the information. At this time, rendering processing such as alpha blending is performed by using Equation 6 based on each component of Y, U, V, and A instead of R, G, B, and A. Graphics processor 14
The YUVA value calculated by c is stored in the corresponding location in the frame buffer 15c determined from the pixel coordinate value. The decoder 16 synchronizes the YUV value stored in the frame buffer 15c with the CRT 18 in synchronization with the scanning of the CRT 18.
Are restored to RGB values for display. Then, this value is converted into analog information by the D / A converter 17 and sent to the CRT 18. The CRT 18 displays the analog information received from the D / A converter on the display.

Here, similarly to the case where one U and V value is assigned to four pixels, one A value is also assigned to four pixels. That is, as shown in FIG. 8, the display of FIG.
Pixels p0 (100), p1 (101), p2 (1
02) and p3 (103) have transparency A0 (400) and pixel p4 (10)
4), p5 (105), p6 (106), p7 (107) have transparency A4
(404) is assigned.

FIG. 13 shows a flow chart of alpha blending by graphics when one A value is given to four pixels, and the flow will be described below.

First, the coordinates of the vertices of the polygon to be drawn and the YUVA value are acquired (step 70). next,
It is judged whether or not all the YUV values of the pixels forming the polygon have been calculated, and if the calculation is completed, the process ends (step 71). Otherwise, the x coordinate of the pixel and YUV
The A value is saved (step 72), 2 is added to the y coordinate every time the calculation for two lines is completed, and the x coordinate and YU are added.
The increment value is added to the VA value (step 76). If the calculation of the YUVA values of the pixels for two lines has not been completed (step 73), rendering processing such as alpha blending using equation 6 is executed (step 74), x
Add 2 to coordinates and increment value to YUVA value (step
75) The process is repeated until the calculation for two lines is completed. The rendering process includes a process of storing the YUVA value being calculated in the memory.

As a result, for example, when performing alpha blending, in the conventional method, the equation (5) was executed once for each pixel of R, G, B, A for each pixel. However, according to the present invention, four pixels have four pixels for Y.
Since it is sufficient to execute once for each of the U, V, and A components, the calculation amount is reduced to about 7/16. Also, the memory capacity and the transfer amount to the frame buffer are reduced to 7/16.

FIG. 14 is a block diagram of another embodiment according to the present invention. The feature of this embodiment is that it is a hidden surface when displaying a three-dimensional object in addition to the luminance value, color difference value and alpha value of the pixel. Z, the parameter used to process the pixel depth
The value has been added.

The present configuration has a computer 11d that owns graphic data 12 including graphic information such as vertex coordinates of polygons and color information indicating the intensity of RGBA values, and a rendering device that calculates the color of the pixels that make up the polygon. 20 and YU
It has a decoder 16, a D / A converter 17 and a CRT 18 for converting V values into RGB values.

Further, the rendering device 20 uses the encoder 13 for converting the RGB values into the YUV values and the YUVA values of the vertices of the polygon from the YU of all the pixels constituting the polygon.
The graphics processor 14d for calculating the VA value, the frame buffer 15d for storing the YUVA value of each pixel, and the Z buffer 19 for storing the Z value of each pixel are provided.

First, the Z-buffer test, which is one of the three-dimensional graphics techniques, will be described.

The Z-buffer test is mainly used for hidden surface processing, and requires a Z-buffer that stores the distance (Z value) between the viewpoint and the object including the pixel for each pixel. The specific operation will be described. For each pixel, the Z value calculated in the rendering process is compared with the value stored in the Z buffer, and the Z value during the rendering process is stored in the Z buffer. If it is smaller than the value (the Z buffer test is satisfied), the YUVA value of the pixel is written in the frame buffer, and the Z value during the rendering process is written in the Z buffer.

Next, the operation of the rendering system having the system configuration shown above will be described.

First, the computer 11d outputs the RGB value of the given figure data 12 to the encoder 13 and the figure information including the A value and the three-dimensional coordinates to the graphics processor 14d in order to display the three-dimensional figure. To do. The encoder 13 that has acquired the RGB value converts the RGB value into the YUV value,
The result is output to the graphics processor 14d.
The graphics processor 14d which has received the YUVA value and the three-dimensional coordinate value calculates the YUVA value and the coordinate value of all the pixels forming the polygon based on the information. At that time, the rendering method is applied based on each component of Y, U, V, and A instead of R, G, B, and A. When the Z buffer test is valid, the Z value being rendered by the graphics processor 14d is compared with the value stored in the corresponding position in the Z buffer 19 determined from the coordinate value of the pixel. Then, only when the comparison operation is established, the YUVA value calculated by the graphics processor 14d is determined from the pixel coordinate value in the frame buffer 15c.
, And the Z value is stored in the corresponding portion of the Z buffer. The decoder 16 synchronizes with the scanning of the CRT 18,
CR the YUV value stored in the frame buffer 15d
Restore RGB values for display at T18. Then, this value is converted into analog information by the D / A converter 17 and sent to the CRT 18. The CRT 18 displays the analog information received from the D / A converter on the display.

Here, as in the case where one U, V and A value is assigned to four pixels, one Z value is assigned to four pixels. That is, with respect to the display image of FIG.
As shown in FIG. 15, pixels p0 (100), p1 (101), p2
In (102) and p3 (103), Z0 (500) corresponds to pixel p4 (104) and p
Z4 (504) corresponds to 5 (105), p6 (106), and p7 (107).

FIG. 16 shows the shading by the graphics when the Z value is set to one for every four pixels, and the alpha.
The flowchart of the blending and Z buffer test is shown, and the flow will be described below.

First, the coordinates of the vertices of the polygon to be drawn and the YUVA value are acquired (step 80). Next, it is judged whether or not all the YUV values of the pixels forming the polygon have been calculated, and if the calculation is completed, the process ends (step
81). Otherwise, the x and z coordinates of the pixel, Y
The UVA value is saved (step 82), every time the calculation of the YUVA value of the pixels for two lines is completed, 2 is added to the y coordinate and the increment value is added to the x coordinate, the z coordinate and the YUVA value ( Step 86). If the YUV of pixels for 2 lines
If the calculation of the A value is not completed (step 83), rendering processing such as alpha blending and Z buffer test is executed (step 84). The Z-buffer test is performed once for 4 pixels. Then, 2 is added to the x coordinate and the increment value is added to the z coordinate and the YUV value (step 85). Steps 84 and 85 are repeatedly executed until the calculation for two lines is completed.

As a result, for example, when the Z buffer test is executed, the Z in the rendering process is changed by the conventional method.
According to the present invention, it is necessary to compare the value with the value stored in the Z buffer once for each pixel.
Since it needs to be executed once for each pixel, the calculation amount of the Z buffer test is reduced to about 1/4. Also, if the Z value is 32,
The size of the screen displayed in bits is 640 x 32.
Assuming that the number of pixels is 0, conventionally, 640 × 320 × 32 bits = about 800 kilobytes was required for the Z buffer memory, but according to the present invention, about 200 kilobytes is sufficient. Further, the amount of data transferred to the Z buffer is reduced to 1/4, and the drawing performance is improved.

FIG. 17 shows the case where fog processing, which is one of the three-dimensional graphics techniques, is executed according to the present invention.
The fog mixing coefficient assigned to each pixel is shown.

The fog processing is indispensable for realizing a simulation such as a flight simulator in which the field of view is not clear. The farther the object to be displayed is from the viewpoint, the greater the degree of fog. Assuming that the fog colors are Yf, Uf, Vf, and Af, the pixel color is calculated as follows based on the fog mixing coefficient f indicating the degree of fog.

[0083]

(7) Y = Y * f + Yf * (1−f) U = U * f + Uf * (1−f) V = V * f + Vf * (1−f) A = A * f + Af * (1−f) ( (7) In fog processing, as shown in FIG. 17, pixels p0 (100), p1 (101), and
In p2 (102) and p3 (103), the fog mixing coefficient f0 (60
0), pixels p4 (104), p5 (105), p6 (106), p7 (107)
Corresponds to the fog mixing coefficient f4 (604), which corresponds to 1 for 4 pixels.
One fog mix coefficient is assigned.

FIG. 18 shows that the fog mixing coefficient f value is 1 for every 4 pixels.
A flowchart of a rendering process including shading by graphics, alpha blending, Z buffer test, and fog in the case where individual pieces are provided is shown. The flow will be described below.

First, the coordinates of the vertices of the polygon to be drawn and the YUV value are acquired (step 90). Next, it is judged whether or not all the YUV values of the pixels forming the polygon have been calculated, and if the calculation is completed, the process ends (step 91). Otherwise, the x and z coordinates of the pixel,
The YUVA value and the f value are saved (step 92), and 2 is added to the y coordinate every time the calculation of the YUVA value of the pixels for two lines is completed, and the x coordinate, the z coordinate, the YUVA value, and the f value are added. The increment value is added (step 96). If the calculation of the YUVA values of the pixels for two lines has not been completed (step 93), rendering processing such as alpha blending, Z buffer test, fog, etc. is executed (step 94). In the fog processing, first, the fog mixing coefficient f is calculated for every four pixels, and the YUV is calculated using Equation 7 based on the f.
Obtain the A value. The fog mixing coefficient f is calculated only once for four pixels. Then, 2 is added to the x coordinate, and the increment value is added to the z coordinate, f value, and YUV value (step
95). Steps 94 and 95 are repeatedly executed until the calculation for two lines is completed.

As a result, when performing fog processing, in the conventional method, it was necessary to execute the equation (7) for each pixel in order to obtain the fog mixing coefficient f. Since it is only necessary to execute once for each pixel, the amount of calculation for obtaining the fog mixing coefficient is reduced to 1/4.

FIG. 19 is a block diagram of another embodiment of the present invention. The feature of this embodiment is that the YUV value stored in the frame buffer is output to the CRT without passing through the decoder.

This configuration has a computer 11d that owns graphic data 12 including graphic information such as polygon coordinates and color information indicating the intensity of RGBA values, and a rendering device 20d that calculates the color of the pixels that form the polygon. , D / A
It has a converter 17 and an NTSC CRT 18d.

Further, the rendering device 20d uses the encoder 13 for converting the RGB values into YUV values and the YUVA values of the vertices of the polygon to determine the Y of all the pixels forming the polygon.
The graphics processor 14d for calculating the UVA value, the frame buffer 15d for storing the YUVA value of each pixel, and the Z buffer 19 for storing the Z value of each pixel are provided.

Next, the operation of the rendering device having such a device configuration will be described.

First, the computer 11d outputs the RGB values of the given figure data 12 to the encoder 13 and the figure information including the A value and the three-dimensional coordinates to the graphics processor 14c in order to display the three-dimensional figure. To do. The encoder 13 that has acquired the RGB value converts the RGB value into the YUV value,
The result is output to the graphics processor 14c.
The graphics processor 14c, which has received the YUVA value and the three-dimensional coordinates, calculates the YUVA value and the coordinate value of all the pixels forming the polygon based on the information. Graphics processor 14c if Z-buffer test is enabled
The Z value calculated in step 1 is compared with the value stored in the corresponding portion of the Z buffer 19 determined from the coordinate value of the pixel. Then, only when the comparison operation is established, the YUVA value calculated by the graphics processor 14c is determined from the pixel coordinate value in the frame buffer 15c.
The calculated Z value is stored in the corresponding portion of the Z buffer. The D / A converter 16 is a CRT 18
The YUV value stored in the frame buffer 15c is converted into analog information and output to the CRT 18 in synchronism with the scanning of. CRT18 is YU received from D / A converter
The analog information of V value is displayed on the display.

When outputting an image to an NTSC CRT, it is sufficient to input the YUV value, so that the decoder can be removed and the total hardware amount can be reduced. In addition, the information in the frame buffer is directly transferred to C without passing through the decoder.
Since it can be output to the RT, the drawing performance until the image is output to the screen is improved.

[0093]

When the conventional rendering process and the rendering process according to the present invention are compared, the color component of the pixel is R,
In the case of G and B, conventionally, it was necessary to calculate each component of R, G, and B for each pixel, but according to the present invention,
By converting R, G, and B into Y, U, and V, the Y component is calculated for each pixel, and the U and V components are calculated for every four pixels. , The capacity of the frame buffer is also reduced by half.

When the color components of the pixels are R, G, B and A, according to the present invention, the Y component of each pixel and the U, V and A components of every 4 pixels are selected. Since the calculation is performed, the calculation amount is reduced to about 7/12 and the capacity of the frame buffer is reduced to 7/12.

Also, when the Z value indicating the depth of a pixel is calculated, the calculation amount of the Z buffer test is reduced to 1/4 by calculating one representative value for every 4 pixels, and the capacity of the Z buffer is reduced. Is also reduced to 1/4.

Further, by having one fog mixing coefficient for every four pixels, the processing for calculating the fog mixing coefficient becomes about 1/4.

Further, in the texture mapping, which is one of the graphics techniques, the process of pasting moving image data such as MPEG in which the data consisting of YUV values is compressed onto polygons can be efficiently performed without requiring any special conversion. Can be executed.

[Brief description of the drawings]

FIG. 1 is a block diagram of a system that is an embodiment of the present invention.

FIG. 2 is a block diagram of a conventional system.

FIG. 3 is a flowchart showing the processing contents of conventional rendering processing.

FIG. 4 is a flowchart showing the processing contents of rendering processing to which the present invention is applied.

FIG. 5 is an explanatory diagram of pixels corresponding to an image on a display.

FIG. 6 is an explanatory diagram of a brightness value assigned to each pixel.

FIG. 7 is an explanatory diagram of color difference values assigned to each pixel.

FIG. 8 is an explanatory diagram of a triangle shading process.

FIG. 9 is a flowchart showing the processing contents of rendering processing according to an embodiment in which the spatial resolution of color difference values is not reduced.

FIG. 10 is a flowchart showing the processing content of rendering processing according to an embodiment in which the spatial resolution of color difference values is reduced.

FIG. 11 is a block diagram of a system according to an embodiment with added transparency.

FIG. 12 is an explanatory diagram of an alpha value assigned to each pixel.

FIG. 13 is a flowchart showing the processing contents of rendering processing according to an embodiment in which the spatial resolution of transparency is lowered.

FIG. 14 is a block diagram of a system according to an embodiment including depth information.

FIG. 15 is an explanatory diagram of a Z value assigned to each pixel.

FIG. 16 is a flowchart showing the processing contents of rendering processing according to the embodiment in which the spatial resolution of the Z value is reduced.

FIG. 17 is an explanatory diagram of fog mixing coefficients assigned to each pixel.

FIG. 18 is a flowchart showing the processing contents of rendering processing according to an embodiment in which the spatial resolution of fog mixing coefficient is reduced.

FIG. 19 is a block diagram of a system according to another embodiment.

[Explanation of symbols]

 11 ... Computer, 12 ... Graphic data, 13 ... Encoder, 14 ... Graphics processor, 15 ... Frame buffer, 16 ... Decoder, 17 ... D / A converter, 18 ... CRT, 20 ... Rendering device.

Claims (8)

[Claims] 1. A graphics rendering device for calculating color information of internal pixels of a polygon called a polygon which constitutes an object to be drawn, wherein each intensity of R, G, B of a pixel is represented by luminance information. And the conversion means for converting into color difference information, and the brightness information of the pixels forming the polygon from the brightness information and the color difference information of each vertex of the polygon,
The conversion means includes a calculation means for calculating color difference information and a storage means for storing the luminance information and the color difference information, and the conversion means calculates each intensity of R, G, B given as color information of each vertex of the polygon. , The luminance information and the color difference information are converted, and the calculating means calculates the luminance information and the color difference information of the pixels forming the polygon based on the luminance information and the color difference information of each vertex of the polygon acquired from the converting means. ,
A rendering device for performing processing of storing the luminance information and the color difference information in the storage means. 2. A conversion device for converting the respective intensities of R, G, B of a pixel into luminance information and color difference information, and color information of each vertex of a polygon called a polygon, from the color information of a pixel inside the polygon. In a rendering method of graphics for calculating color information of pixels constituting a polygon in a rendering device having a calculation device for calculating the color information and a storage device for storing the color information, each vertex of the polygon is provided with the color information as color information. R,
The respective intensities of G and B are converted into luminance information and color difference information using the conversion means, and based on the luminance information and color difference information of each vertex of the polygon, the luminance information and color difference information of the pixels inside the polygon. Is calculated using the arithmetic unit, and the luminance information and the color difference information are stored in the storage device. 3. The color computing device according to claim 1, wherein the computing means uses the color difference information of one pixel for a plurality of pixels as a representative of the plurality of color difference information to calculate color difference information of pixels forming the polygon. A rendering device that calculates and stores the color difference information in the storage unit as a representative of the plurality of color difference information. 4. The calculation means according to claim 1 or 3, wherein, in order to calculate transparency information indicating transparency of a pixel to be drawn, the calculation means outputs transparency information of one pixel to the plurality of pixels. As a representative of the transparency information, a graphic for calculating the transparency information of the pixels forming the polygon based on the transparency information of each vertex of the polygon and storing the transparency information in the storage means as a representative of the plurality of transparency information. Rendering device. 5. The drawing means for drawing three-dimensional graphics according to claim 1, 3 or 4, wherein the computing means uses depth information of one pixel as a representative of the plurality of pixels for the plurality of pixels. A graphics rendering in which the depth information of pixels forming the polygon is calculated based on the depth information of each vertex of the polygon, and the depth information is stored in the storage unit as a representative of the plurality of depth information. apparatus. 6. The method according to claim 1, 3, 4 or 5, for calculating fog information representing a degree of fog derived from a distance between a viewpoint and an object, which is used to generate a fog effect on the entire screen. The computing means uses fog information of one pixel for a plurality of pixels as a representative of the plurality of fog information, and based on the fog information of each vertex of the polygon,
A graphics rendering device for calculating the fog information of pixels constituting the polygon. 7. A control device comprising the rendering device according to claim 1, 3, 4, 5 or 6, and having a geometrical information and color information of a shape to be displayed, the luminance information and the color difference information being R, A second converter for converting into G and B intensities and a display device for displaying the rendering processing result are provided, and the rendering device acquires geometric information and color information of the shape to be displayed from the control device. , The luminance information, the color difference information, the transparency information, and the depth information of the pixels inside the shape to be displayed are stored in the storage means, and the second converter stores the luminance information and the color difference information stored in the storage means as R and G. , B intensity,
A rendering system for performing processing of displaying on the display means. 8. A control device including the rendering device according to claim 1, 3, 4, 5, or 6, and a control device having geometric information and color information of a shape to be displayed, and a display for displaying a rendering processing result. The rendering device acquires geometric information and color information of a shape to be displayed from a control device, and stores brightness information, color difference information, transparency information, and depth information of pixels inside the shape to be displayed. Memorize in the means,
A rendering system in which the display unit performs a process of displaying a shape based on the luminance information and the color difference information stored in the storage unit.