JP2001126084A - Device and method for displaying texture and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress aliases when a texture is reduced and displayed and to reduce the capacity of the data quantity of texture data. SOLUTION: This texture display device is provided with a storing part 36 for orthogonally transforming each spatial frequency component and storing quantized texture data, inversely quantizing parts 42 and 43 for setting-the value of an inversely quantization coefficient equivalent to the high frequency component of the texture data to be a small value in accordance with the reduction ratio of the texture when the texture is displayed to execute inverse quantization so as to attenuate the high frequency component, an inverse transformation part 44 for inversely transforming the texture data and display parts 33, 35 and 38 for displaying the texture by desired supplement processing.

Description

DETAILED DESCRIPTION OF THE INVENTION

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image display technique, and more particularly, to a technique for displaying a texture in a reduced size.

2. Description of the Related Art In the field of computer graphics, techniques for reducing and displaying a texture on a screen are disclosed, for example, in JP-A-11-250280 and JP-A-11-11-1.
No. 75739, JP-A-10-307925,
As disclosed in JP-A-9-231383, a technique using a mipmap texture is known. The mipmap texture is a texture created by reducing the size from the same size to the size of one pixel of the original size texture by 1/2 power. According to this technology,
Store the mipmap texture in memory in advance,
When mapping a texture, a mipmap texture with an appropriate reduction ratio is selected from the ratio between the object size and the texture size, and this is mapped to a polygon.

However, if a mipmap texture is prepared, there is a problem that the amount of texture data increases by 33%. Since the types of textures increase in order to display a high-definition image, it is desirable that the amount of texture data be small in consideration of a limited memory space. On the other hand, as a technique for reducing and displaying a texture on a screen, in addition to a technique using a mipmap texture, a technique for enlarging / reducing a texture by performing linear interpolation processing on texture data can be considered. Examples of the linear interpolation processing include nearest neighbor interpolation, bilinear interpolation (linear interpolation between four adjacent points), and cubic spline interpolation (spline interpolation of 16 nearby points). In each of these cases, the data interval of the texture data at the time of enlargement or reduction is calculated with respect to the original data interval, and based on this, the position of the interpolation point in the original texture is calculated. Is determined, and the value at that point is determined by various interpolation methods. However, in the linear interpolation processing, there is a problem that aliasing occurs when the texture is reduced and displayed. The cause is that when the frequency component higher than the Nyquist frequency determined by the sampling frequency at the time of data complementation is included in the texture data, the frequency component higher than the Nyquist frequency is turned back and becomes a spectrum that causes an alias (false image). That's why. As a countermeasure in such a case, it is conceivable to reduce the sampling frequency. However, if there is a limitation on hardware, this method cannot be adopted. As another method, it is conceivable to reduce aliasing of the spectrum by passing through a low-pass filter to reduce aliasing. However, this method requires a large-scale two-dimensional low-pass filter circuit, and also requires a texture data. However, there is a problem that the effective component of this may also be removed. Accordingly, it is an object of the present invention to suppress aliasing when a texture is reduced and displayed and to reduce the amount of texture data.

In order to solve the above problems, the present invention provides an orthogonal transform (encoding) for each spatial frequency component.
Prepared texture data is prepared in advance, and when displaying the texture, after attenuating high frequency components of the texture data in accordance with the reduction ratio of the texture, the texture data is inversely transformed (decoded), and The texture is displayed by the complement processing of. According to such a procedure, the high-frequency components included in the texture data are previously attenuated when the texture is reduced and displayed, so that the occurrence of aliases can be suppressed. Further, according to the present invention, since it is not necessary to prepare a mipmap texture, it is possible to reduce the amount of texture data. Further, according to another aspect of the present invention, orthogonally transformed for each spatial frequency component, and further, quantized texture data is prepared in advance, and when displaying the texture, the inverse data corresponding to the high frequency component of the texture data is displayed. By setting the value of the quantization coefficient to a small value corresponding to the reduction ratio of the texture, inverse quantization is performed so as to attenuate the high frequency component, and after the texture data is inversely transformed, a desired interpolation process is performed. To display the texture. In the present invention, the XY directions orthogonal to each other with respect to the texture are defined, and the high frequency of the texture data in the X direction and the Y direction are independently corresponding to the reduction ratio in the X direction and the reduction ratio in the Y direction. It is possible to attenuate the components. By such a method, it is possible to realize an appropriate reduced display according to the reduction ratio in the X direction and the Y direction. Also, regarding the bit assignment of texture data orthogonally transformed for each spatial frequency component, a longer bit length is assigned to texture data as the inverse quantization coefficient is smaller, and a shorter bit length is gradually assigned as the inverse quantization coefficient increases. By reducing the data amount of the high frequency component, the data amount of the entire texture data can be reduced. Further, according to the present invention, it is possible to provide a computer-readable recording medium in which a program for executing the above procedure is recorded. Here, the recording medium is a medium on which an image processing program or the like is recorded by some physical means, and is a computer, particularly,
A processor that can realize a desired function by a dedicated processor (for example, a video display processor). Therefore, any device may be used as long as it can be downloaded to the computer by some means to realize a desired function. For example, ROM, flexible disk, hard disk, CD-ROM, CD-R, DVD-ROM, DVD
-RAM, DVD-R, PD disk, MD disk,
Includes information recording media such as MO disks. This includes the case where data is transferred from the host computer via a wired or wireless communication line (public line, data line, satellite line, etc.). The so-called Internet is also included in the recording medium mentioned here. It should be noted that in the present specification, “...” is a concept realized in the processing procedure of the texture data, and does not always correspond one-to-one to a specific hardware or software routine. In some cases, the same hardware element implements a plurality of "... parts", or one "..." in relation to a plurality of hardware elements.
Unit "may be realized.

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows a block diagram of the game device. As shown in FIG. 1, the game device mainly includes a CPU block 10, a main bus 20, a VDP (video display processor) 30, a texture memory 36, a frame memory 37, and a TV monitor 38. The CPU block 10 includes a main CPU 11, an RO
M12, RAM13, pad interface (pad I
/ F) 14, control pad 15, CD-ROM interface (CD-ROM I / F) 16, and CD
-It is provided with a ROM 17. The CPU block 10 controls the execution of the game program in response to the player's input operation, and further specifies the reduction ratio of the texture displayed on the TV monitor 38 to the VDP 30. Texture data for mapping to game programs and objects, parameters for specifying polygon shapes and their textures and the like are stored in the CD-ROM 17, and the programs and parameters are stored in the CD-ROM I / O.
The data is transferred to the work area of the RAM 12 via / F16. The main CPU 11 executes the program transferred to the RAM 13 in response to the input signal of the control pad 15 operated by the player, and appropriately designates a polygon drawing command to the VDP 30 based on the parameters. The texture data stored in the CD-ROM 17 is transferred to the texture memory 36 via the main bus 20, main bus I / F 31, and texture processing unit 32. Texture memory 3 from CD-ROM 17
6 has been subjected to DCT (discrete cosine transform) and quantization processing in advance. As shown in FIG. 3, the bit assignment of the texture data stored in the texture memory 36 is 8 bits for the DC component (upper left in the figure) and 7 bits from the low frequency component to the high frequency component (lower right in the figure). , 6 bits,..., 2 bits, and 1 bit. As described above, by reducing the bit assignment of the high-frequency component, it is possible to previously reduce the amount of unnecessary high-frequency component data in the reduced display of the texture. Data volume is 47
% Can be reduced. Of course, the way of bit assignment is not limited to the pattern shown in the figure, and may be assigned according to the quantization coefficient. The VDP 30 includes a main bus interface (main bus I / F) 31, a texture processing unit 32,
It comprises a drawing processing unit 33, a frame memory control unit 34, and a display processing unit 35. VDP30 is TV
A processor that controls image display on the monitor 38.
The polygon drawing command is supplied from the main CPU 11 to the drawing processing unit 33 via the main bus I / F 31.
The texture processing unit 32 performs inverse quantization and inverse DCT processing on the texture data in accordance with the reduction ratio of the texture specified by the rendering processing unit 33 based on the polygon rendering command when displaying the texture image. , And transfers the processing result to the drawing processing unit 33. The drawing processing unit 33 performs complementary processing such as bilinear interpolation to generate texture data of the texture reduced to a desired ratio, and transfers the processing result to the frame memory control unit 34. The frame memory control unit 34 writes the pixel data in the frame buffer 37 based on the texture data, reads the pixel data in synchronization with the image update period, and transfers the pixel data to the display processing unit 35. Display processing unit 3
Reference numeral 5 performs D / A conversion of pixel data, video signal generation processing, and the like, and displays an image on the TV monitor 38. Next, a detailed configuration of the texture processing unit 32 will be described with reference to FIG. The texture processing unit 32 includes a texture memory control unit 41, a spatial frequency limiter 42, an inverse quantizer 43, and an inverse DCT unit 44 therein. The spatial frequency limiter 42 filters the inverse quantization table value of the texture data stored in the texture memory 36 in accordance with the reduction ratio of the texture (the details of the filter process will be described later). The inverse quantization table is a table in which DCT and inverse quantization coefficients for inversely quantizing the quantized texture data are arranged in a matrix (for example, 8 × 8). The texture memory control unit 41 reads out DCT and quantized texture data from the texture memory 36 and transfers the data to the inverse quantizer 43.
The inverse quantizer 43 performs inverse quantization of the texture data based on the filtered inverse quantization table supplied from the spatial frequency limiter 42, and outputs the processing result to the inverse DCT unit 4.
Transfer to 4. The inverse DCT unit 44 performs an inverse DCT on the texture data, and transfers a processing result to the drawing processing unit 33. Next, a detailed configuration of the spatial frequency limiter 42 will be described. The spatial frequency limiter 42 includes an inverse quantization coefficient memory 51 and a high frequency component attenuation calculator 52 inside. In the inverse quantization coefficient memory 51, an inverse quantization table corresponding to each texture data is set as a default value. The high frequency component attenuation calculator 52 reads the inverse quantization table from the inverse quantization coefficient memory 51 and performs a filtering process on the inverse quantization table in accordance with the texture reduction ratio specified by the drawing processing unit 33,
The high-frequency component of the inverse quantization coefficient is attenuated. The high frequency component attenuation calculator 52 generates a filter matrix based on the reduction ratio of the texture, and filters the inverse quantization table. As a result of the filtering, the high-frequency components of the texture data inversely quantized by the inverse quantizer 43 are attenuated, so that it is possible to avoid the inconvenience that aliasing occurs when the drawing processing unit 33 performs linear interpolation.
The inverse quantization table may be set at the time of drawing a texture, or may be set before drawing. Here, the filtering process will be described with reference to FIGS. FIG. 4 is an inverse quantization table registered as a default value in the inverse quantization coefficient memory 51. The (i, j) component (inverse quantization coefficient) located at the i-th row and the j-th column is denoted by Q _ij . In the present embodiment, since processing is performed in units of 8 × 8 pixels, 1
≤ i ≤ 8, 1 ≤ j ≤ 8. FIG. 5 shows a filter matrix, and its (i, j) component is denoted by F _ij . The filter matrix is a matrix for setting an inverse quantization coefficient corresponding to a high frequency component in the inverse quantization table to a smaller value or “0” (attenuation processing), and each F _ij (1
≤ i ≤ 8, 1 ≤ j ≤ 8) satisfies the relationship 0 ≤ F _ij ≤ 1. FIG. 6 shows the inverse quantization table after the filter processing. The (i, j) component is obtained by multiplying the inverse quantization table by the corresponding (i, j) component of the filter matrix. Therefore, its (i, j) component is Q
_ij F _ij . A specific example of the filtering process will be described with reference to FIGS. First, the reduction ratio of the texture is defined. When an X direction (vertical direction) and a Y direction (horizontal direction) orthogonal to each other are set with respect to the texture, LOD = n is set when the texture is reduced and displayed at 2− ⁿ times in both directions. LOD = 0
In other words, as shown in FIG. 7A, when the image is displayed with the texture at the same magnification, it is not necessary to attenuate the high frequency components, and therefore all the components of the filter matrix F
_ij (1 ≦ i ≦ 8, 1 ≦ j ≦ 8) is set to “1”. On the other hand, when LOD = 1, that is, as shown in FIG. 3B, when displaying an image with the texture being 1/2 times the X direction and the Y direction, F _8j (1 ≦ j ≦ 8) ) And F
_i8 (1 ≦ i ≦ 7) is set to “0”, and the other components are set to “1”. When LOD = 2, that is, as shown in FIG.
To display an image at 4 times, F _7j (1 ≦ j ≦
8), F _8j (1 ≦ j ≦ 8), F _i7 (1 ≦ i ≦ 6),
And F _i8 (1 ≦ i ≦ 6) are set to “0”, and the other components are set to “1”. As described above, by setting each component of the filter matrix corresponding to the high-frequency component of the inverse quantization table to “0” corresponding to the LOD value, that is, by cutting the high-frequency component of the texture data, the drawing process is performed. Part 33
Since a component having a frequency higher than the Nyquist frequency determined by the sampling frequency can be cut in advance at the time of the complementing process, the occurrence of an alias can be prevented. In the present invention, the LOD value can be applied not only to integers but also to positive numbers including decimal numbers.
For example, if the texture is 2 in the X and Y directions
^-When displaying an image at ^0.5 times, that is, LOD =
In the case of 0.5, as shown in FIG. 8A, F _8j (1
≦ j ≦ 8) and F _i8 (1 ≦ i ≦ 7) are set to “0.5”, and the other components are set to “1”. Also, when displaying an image with the texture being ^2-1.3 times in the X and Y directions,
That is, when LOD = 1.3, as shown in FIG. 8B, F _8j (1 ≦ j ≦ 8) and F _i8 (1 ≦ i ≦ 7) are set to “0”, and F _7j (1 ≦ j ≦ 7) and F _i7 (1 ≦ i
≦ 6) is set to “0.7”, and the other components are set to “1”. Here, generalizing the expression for obtaining the (i, j) component of the filter matrix for the case of displaying an image at 2- ⁿ times the X direction and the Y direction, the following expression is obtained. F _ij = min ｛S (1- (n- (8-i))), S
(1- (n- (8-j)))} Here, the function S (x) takes a value of “0” when x ≦ 0, and takes a value of “x” when 0 <x <1. A value is taken, and a value of “1” is taken when 1 ≦ x. Also, the function min
{X, y} takes the value of “x” when x <y, and y ≦ x
In this case, the value of "y" is taken. As described above, according to the present invention, since a texture display of an arbitrary size is possible, when there is no mipmap texture having an appropriate reduction ratio as in the related art, a plurality of approximate mipmap textures are selected. There is no need to perform linear interpolation.
Further, in the present invention, the spatial frequencies in the X direction and the Y direction can be independently controlled. For example, X
When the texture is reduced by 2- ⁿ times in the direction, LODX = n is defined, and when the texture data is reduced by 2- ^m in the Y direction, LODY = m is defined. For example, when LODX = 1 and LODY = 0, that is, when the texture is displayed in 倍 times only in the X direction, as shown in FIG. 9A, F _i8 (1 ≦ i ≦
8) is set to "0" and the other components are set to "1". Also, L
When ODX = 0 and LODY = 1, that is, when the texture is displayed in 倍 times only in the Y direction, F _8j (1 ≦ j ≦ 8) is set to “0” as shown in FIG. "And the other components to" 1 ". Also, LODX = 1, LODY =
2, ie, the texture is １ ／ with respect to the X direction.
When the image is displayed at 1/4 times the magnification in the Y direction, as shown in FIG. 4C, F _7j (1 ≦ j ≦ 8), F _7j
8j (1 ≦ j ≦ 8) and F _i8 (1 ≦ i ≦ 6) are set to “0”, and the other components are set to “1”. Thus, L
By limiting the spatial frequency components of the texture data in the X direction and the Y direction independently according to the ODX value and the LODY value, it is possible to attenuate the high frequency components suitable for the respective reduction ratios. Conventionally, a method called a lip map (LIPMAP) is known as a method of using a texture in which the reduction ratio is independently changed in the X direction and the Y direction. However, in such a method, the maximum texture data amount is reduced. 300% of the texture data amount is required. On the other hand, according to the present invention, the data amount can be reduced by about 87% compared to the lip map method. In the present invention, the LODX value and the LODY value can be applied not only to integers but also to positive numbers including decimal numbers. For example, LO
DX = 0.5, when lody = 1.3, i.e., ^{2 -0.5} times the texture with respect to the X direction, the Y direction 2
When displaying at ^-1.3 times, as shown in FIG. 10A, F _8j (1 ≦ j ≦ 8) is set to “0”, and F _7j (1 ≦
j ≦ 7) is set to “0.7”, F _i8 (1 ≦ i ≦ 7) is set to “0.5”, and the other components are set to “1”. in this case,
F ₈₈ is referring to the value of LODX = 0.5 "0.5"
And becomes “0” when the value of LODY = 1.3 is referred to, but in this example, a smaller value, that is, “0” is adopted. Similarly, F ₇₈ becomes “0.5” when referring to the value of LODX = 0.5, and becomes “0.7” when referring to the value of LODY = 1.3. To “0.5”. However, when the value of the specific component of the filter coefficient is different from the value obtained from the LODX value and the value obtained from the LODY value, in addition to using the smaller value as described above, the average value of the two and the value Or the square root of the product of. LODX = 2.4, L
When ODY = 1.3, i.e., ^{2 -2.4} times the texture with respect to the X direction, when viewed in ^{2 -1.3} times the Y direction, as shown in (B) , F _8j (1 ≦ j
≦ 8) is set to “0”, and F _7j (1 ≦ j ≦ 5) is set to “0.
7 ", and F _i8 (1 ≦ i ≦ 7) and F _i7 (1 ≦ i ≦
7) is set to “0” and F _i6 (1 ≦ i ≦ 7) is set to “0.6”.
And the other components to “1”. In this case, F ₈₆ is L
If the value of ODX = 2.4 is referred to, it becomes “0.6” and L
When the value of ODY = 1.3 is referred to, the value becomes “0”. In this example, a smaller value, that is, “0” is adopted. Similarly, F ₇₆ becomes “0.6” by referring to the value of LODX = 2.4, and becomes “0.7” by referring to the value of LODY = 1.3, but is smaller in this example. Value, ie
“0.6” shall be adopted. Here, LODX =
When the equation for calculating the (i, j) component of the filter matrix is generalized for the case where n, LODY = m, the following equation is obtained. F _ij = min ｛S (1- (n- (8-i))), S
(1- (m- (8-j)))} As described above, according to the present embodiment, when the texture is reduced and displayed, the Nyquist frequency component equal to or higher than the Nyquist frequency component determined by the sampling frequency of the linear interpolation processing or the like is used. Aliasing can be prevented by attenuating high frequency components in advance by filter processing. Further, according to the present embodiment, since it is not necessary to prepare a mipmap texture, the data amount of texture data can be reduced as compared with the conventional method using the mipmap method.
Also, according to the present embodiment, the texture X direction and Y direction
Since the spatial frequency component can be controlled independently for each direction, for example, even if the display is reduced only in the X direction, X
High frequency components in only the direction can be attenuated by filtering. Also, the DC is stored in the texture memory 36 in advance.
T and the quantized texture data are stored, and the inverse DCT processing of the texture data at the time of drawing an image can include a filtering process, so that the calculation load of the VDP can be reduced. Further, by assigning a short bit length to the bit assignment of the texture data stored in the texture memory 36 according to the inverse quantization coefficient, the data amount of the texture data stored in the texture memory 36 can be reduced. This is because when the texture is reduced and displayed, the high-frequency components are attenuated by the filter processing in accordance with the reduction ratio, so that the data amount allocated to the high-frequency components in advance is reduced. When color data is used as the texture data, the ratio of the sampling frequency of the luminance information Y and the sampling frequency of the color difference information Cb, Cr is set to, for example, 4: 2: 2, 4: 1: 1, or the like. The compression ratio can be increased. In this case, an inverse quantization table for the luminance signal and an inverse quantization table for the chrominance signal may be prepared, and each of them may be subjected to the filter processing according to the reduction ratio of the texture. Further, in the above description, discrete cosine transform has been described as an example of orthogonally transforming texture data for each spatial frequency component, but discrete Fourier transform (DFT), wavelet transform, or the like may be used. In this specification, texture data means not only image data representing a pattern rendered on the surface of a polygon, but also any image drawn on a screen (for example, a foreground image of a character or an object, or a background image drawn on a background). Background image, etc.).

According to the present invention, when a texture is reduced and displayed, aliasing can be prevented by attenuating high frequency components of the texture data in advance in accordance with the reduction ratio. Further, according to the present invention, it is not necessary to prepare a mipmap texture at the time of the reduced display of the texture, so that the amount of texture data required for the reduced display can be reduced. Further, according to the present invention, since the spatial frequency component of the texture data can be controlled independently for the X direction and the Y direction of the texture, an appropriate reduced display of the texture according to the reduction ratio in the X direction and the Y direction can be achieved. Can be realized.

[Brief description of the drawings]

FIG. 1 is a block diagram of a game device.

FIG. 2 is a block diagram of a texture processing unit.

FIG. 3 is an explanatory diagram of bit assignment of texture data.

FIG. 4 is an explanatory diagram of an inverse quantization table before filter processing.

FIG. 5 is an explanatory diagram of a filter matrix.

FIG. 6 is an explanatory diagram of an inverse quantization table after filter processing.

FIG. 7 is an explanatory diagram of a filtering process.

FIG. 8 is a specific example of a filter matrix.

FIG. 9 is an explanatory diagram of a filtering process.

FIG. 10 is a specific example of a filter matrix.

[Explanation of symbols]

30 VDP, 31 Main bus I / F, 32 Texture processor, 33 Drawing processor, 34 Frame memory controller, 35 Display processor, 36 Texture memory,
37 frame memory, 38 TV monitor, 41 texture memory control unit, 42 spatial frequency limiter, 43
Inverse quantizer, 44: inverse DCT unit, 51: inverse quantization coefficient memory, 52: high frequency component attenuation calculator

Claims (13)

[Claims] A storage unit for storing texture data orthogonally transformed for each spatial frequency component; reading out the texture data from the storage unit when displaying the texture; A texture display device, comprising: an attenuation processing unit that attenuates a component, an inverse transformation unit that inversely transforms texture data in which a high-frequency component has been attenuated, and a display unit that displays a texture by a desired complementation process. 2. The attenuation processing section defines XY directions orthogonal to each other with respect to the texture, and independently controls the textures in the X and Y directions in accordance with the reduction ratio of the texture in the X direction and the reduction ratio in the Y direction, respectively. The texture display device according to claim 1, wherein the high frequency component of the data is attenuated. 3. The texture display device according to claim 1, wherein the attenuation processing unit attenuates more high frequency components as the texture reduction ratio increases. 4. A storage unit for storing orthogonally transformed and further quantized texture data for each spatial frequency component, and reads out the texture data from the storage unit when displaying the texture, and corresponds to a high frequency component of the texture data. An inverse quantization unit that performs inverse quantization so as to attenuate the high-frequency component by setting the value of the inverse quantization coefficient to a small value corresponding to the reduction ratio of the texture; A texture display device, comprising: an inverse conversion unit that performs an inverse conversion of data; and a display unit that displays a texture in a desired complementary process. 5. The dequantizing unit defines XY directions orthogonal to each other with respect to the texture, and independently responds to the X-direction and Y-direction reduction ratios of the texture in the X and Y directions, respectively. 5. The texture display device according to claim 4, wherein the high frequency component of the texture data is attenuated. 6. The texture display device according to claim 4, wherein the inverse quantization unit attenuates more high frequency components as the texture reduction ratio increases. 7. The texture data stored in the storage unit has a smaller inverse quantization coefficient, a longer bit length is assigned to the texture data, and a shorter bit length is assigned as the inverse quantization coefficient increases. Item 6
The texture display device according to claim 1. 8. Texture data that has been orthogonally transformed for each spatial frequency component is prepared in advance, and when displaying a texture, after attenuating high frequency components of the texture data in accordance with the texture reduction ratio, A texture display method in which texture data is inversely converted and a texture is displayed by a desired complementary processing. 9. An orthogonal transform for each spatial frequency component and texture data that has been further quantized are prepared in advance, and the value of the inverse quantization coefficient corresponding to the high frequency component of the texture data is displayed when the texture is displayed. Texture display that performs inverse quantization to attenuate the high-frequency component by setting it to a small value corresponding to the reduction ratio of, further reverse-transforms the texture data, and displays the texture with desired complementation processing Method. 10. X orthogonal to the texture
The Y direction is defined, and the texture reduction ratio in the X direction and Y
X direction and Y direction respectively corresponding to the reduction ratio in the direction.
The texture display method according to claim 8, wherein the high frequency component of the texture data in the direction is attenuated. 11. The texture display method according to claim 8, wherein as the texture reduction ratio increases, more high frequency components are attenuated. 12. A longer bit length is assigned to the texture data as the inverse quantization coefficient of the texture data orthogonally transformed for each spatial frequency component is smaller, and a shorter bit length is assigned as the inverse quantization coefficient increases.
The texture display method according to any one of claims 11 to 11. 13. The computer according to claim 8, wherein:
3. A computer-readable recording medium on which a program for executing the method according to any one of 2 is recorded.