KR0176495B1 - Texel mapping acceleration apparatus - Google Patents

Abstract

A texel mapping accelerator is disclosed. The apparatus comprises: first storage means for storing texture elements (hereinafter referred to as texels) necessary for obtaining pixel values constituting the screen, temporary storage means for temporarily storing a predetermined number of texels, and texel necessary for obtaining pixel values First and second address generating means for generating transverse and longitudinal addresses of the first and second addresses, and a predetermined number of texels are input from the first storage means and output to the temporary storage means, and corresponding to the mapping method and the transverse and longitudinal addresses. Calculates the texel address, checks whether the texel corresponding to the texel address exists in the temporary storage means, and matches the corresponding texel input from the temporary storage means or the first texel input from the first storage means according to the checked result. Control means for outputting as a first color value and a second color value necessary for shading for adding visual information to the image itself. A mixing means for inputting and mixing the shading means, the first color value and the second color value, and outputting the mixed result as a third color value finally required for obtaining the pixel value; and a third color input from the mixing means. And a second display means for temporarily storing a value as image data and an image display means for inputting image data stored in the second storage means and displaying the image data on the screen. The speed of texture mapping can be significantly improved. In addition, the number of pins of the processor can be greatly reduced, thereby reducing the printed circuit board cost.

Description

TEXEL MAPPING accelerator

1 is a block diagram of a conventional texel mapping apparatus that performs texture mapping.

2 is an overall block diagram illustrating a texel mapping acceleration apparatus according to the present invention.

3 is a block diagram according to the present invention of the mapping unit shown in FIG.

TECHNICAL FIELD The present invention relates to the field of digital computers, and more particularly, to an apparatus for improving the processing speed of computer graphics.

An important application of computer graphics requires the creation of vivid images of three-dimensional objects, and designers of three-dimensional objects such as cars, aircraft, and other structures can easily implement repetitive design changes before creating the prototype of the object. We want to have a visual representation of the preliminary design. Implementing a computer that delivers realistic images can provide a cheaper, faster, and more efficient way of evaluating preliminary designs than architectural models or prototypes.

In particular, the image-providing computer allows designers to evaluate many alternative designs in a short time. Often the design work itself can only be done within a computer system using CAD (Computer Aided Design) or other engineering workstations. In such cases, the digitized representation of the three-dimensional object has already been used as the basis of the computer generating the image.

Simulation systems are another application of computers that produce three-dimensional images, and such systems utilize images that must appear realistic and dynamically change.

To reproduce this image more realistically, it is based on a surface description called a texture. In computer graphics, a texture generally includes two aspects. The first aspect of the texture is to add a rough shape to the soft side. The second aspect is to add a pattern specifically described to the soft side. After the pattern is added, the face still looks smooth. Adding a pattern to the soft side is a mapping operation called texture mapping. Texture mapping is a very important technology in computer graphics, and many methods have been developed to improve texture mapping.

One such method is described in Real-Time Manipulation of Texture Mapped Surfaces, Oka, et al, Computer Graphics, Volume 21, No. 4, pp181-188, 1987. Also, texture mapping is described in Rogers, David F., Procedural Elements. For Computer graphics, McGraw-Hill, 1985, pp354-363.

FIG. 1 is a block diagram of a conventional texel mapping apparatus that performs texture mapping, and includes a geometry unit 200, a processor unit 202, a texture memory unit 204, a frame buffer unit 206, and a screen unit 208. Consists of

A conventional texel mapping operation through the above-described configuration will be described with reference to FIG. 1.

The geometry unit 200 converts, scales, lights, and emits a three-dimensional object, and the processor unit 202 obtains pixel values constituting an image. Extract the texture values (also referred to as texture elements or texels) stored in the texture memory 204, which is a separate memory that stores the texture image, to obtain pixels, and form the image by the mapping process. This formed image is temporarily stored in a frame buffer 206 composed of video RAM before being output to the screen 208.

Generally, at least one texture value is needed to obtain one pixel value, but more texel values per pixel, i.e. 1,2,4 or 8 texel values per pixel, are required to obtain more sophisticated pixel values. This means that there is an entry and exit between the processor unit 202 and the texture memory unit 204. Therefore, when there is only one data path 210 between the processor 202 and the texture memory 204, access to the texture memory 204 is required up to eight times, so that the processing speed is significantly reduced. In many cases, the cost of hardware increases due to an increase in the number of pins of the processor unit 202 and the complexity of a printed circuit board (PCB).

An object of the present invention is to solve the above problems by bringing the texture value in the texture memory to the extra storage unit provided in the processor in advance, so that TEXEL MAPPING can be performed quickly using only one data path between the processor and the texture memory. An object of the present invention is to provide a texel mapping acceleration device that can be performed within the apparatus.

In order to achieve the above object, a texel mapping acceleration device according to the present invention includes first storage means for storing texture elements (hereinafter, referred to as texels) necessary for obtaining pixel values constituting a screen, and temporarily storing a predetermined number of texels. Temporary storage means, first address generating means for generating a lateral address of the texel necessary for obtaining the pixel value, second address generating means for generating a longitudinal address of the texel necessary for obtaining the pixel value, and Input a predetermined number of texels from the first storage means and output them to the temporary storage means, calculate a texel address corresponding to the mapping method and the transverse address and the slave address, and the texel corresponding to the texel address Check whether there is a storage means, and the corresponding texel inputted from the temporary storage means Control means for outputting the corresponding texel input from the first storage means as a first color value, shading means for generating a second color value for shading to add visual information to the image by itself; Mixing means for inputting and mixing the first color value and the second color value, and outputting the mixed result as a third color value finally required for obtaining the pixel value, and the third input from the mixing means; And second image display means for temporarily storing color values as image data and image display means for inputting and displaying the image data stored in the second storage means on the screen.

2 is a block diagram illustrating the texel mapping accelerator according to the present invention, and includes a first storage unit 400, a mapping unit 402, a mixing unit 404, a shadow unit 406, and a second storage unit. 408 and an image display unit 410.

The operation of the texel mapping acceleration apparatus according to the present invention through the above-described configuration will be described in detail as follows with reference to FIG. 2.

The first storage unit 400 illustrated in FIG. 2 stores texture elements (hereinafter, referred to as texels) necessary to obtain pixel values constituting the screen, and reads the stored texels to the mapping unit 402. Here, the first storage unit 400 functionally corresponds to the texture memory unit 204 shown in FIG. 1. The mapping unit 402 inputs a predetermined number of necessary or predicted texels from the first storage unit 400 in order to reduce the time for receiving the texels from the first storage unit 400 to perform the texel mapping. After temporarily storing, the temporarily stored texels are mapped, and the mapped results are output to the mixing unit 404 as a first color value.

Meanwhile, the shader 406 generates a second color value necessary for shading to add visual information to the image and outputs the second color value to the mixer 404. The mixing unit 404 mixes the first color value output from the mapping unit 402 and the second color value output from the shadowing unit 406 as inputs, and mixes the result of the mixing into a finally required third color value. 2 is output to the storage unit 408.

The second storage unit 408 receives the third color value output from the mixing unit 404 and temporarily stores it as image data. Here, the second storage unit 408 functionally corresponds to the frame buffer unit 206 shown in FIG. The image display unit 410 corresponding to the screen unit 208 shown in FIG. 1 displays image data stored in the second storage unit 408 on the screen.

The configuration and operation according to the present invention of the mapping unit 402 illustrated in FIG. 2 described above will be described with reference to FIG. 3.

FIG. 3 is a detailed block diagram of the mapping unit 402 shown in FIG. 2 according to the present invention, and is a block diagram showing a connection state between the mixing unit 404 and the first storage unit 400. The first address generator 600, the second address generator 602, the temporary storage unit 604 and the control unit 606.

Under the control of the controller 606 shown in FIG. 3, a series of texture elements stored in the first storage unit 400 are temporarily stored in the temporary storage unit 604 shown in FIG. 3. That is, the controller 606 inputs texture elements that are frequently used or are expected to be frequently used from the first storage unit 400, and controls the input texture elements to be stored in the temporary storage unit 604. In addition, when a texel required to obtain a final pixel value of a desired pixel is required by the mixer 404, the controller 606 may determine the width and width of the image obtained by the first and second address generators 600 and 602, respectively. Input the addresses and calculate the texel addresses for the required texels according to the mapping methods such as point sampling, bi-linear sampling, and tri-linear sampling, and then correspond to the texel addresses. It is checked whether the texel is stored in the temporary storage unit 604. If the corresponding texel is stored in the temporary storage unit 604, the inputted texel is input to the mixing unit 404 as a first color value so that the final pixel value can be obtained. However, when the corresponding texel is not stored in the temporary storage unit 604, a texel corresponding to the texel address is input from the first storage unit 400 and output as a first color value to the mixing unit 404. The corresponding texel input from the first storage unit 400 may be stored in the temporary storage unit 604.

As a result, as described above, the texel mapping acceleration apparatus according to the present invention is a texture memory that is the first storage unit 400 each time to improve performance in performing texture mapping, which is an important function in a high performance 3D graphics system, using hardware. By accessing the texel value, which is necessary or anticipated, into the texel cache, which is a temporary storage unit 604, in advance, the speed of texture mapping can be significantly improved, and the processor ( By eliminating the need for multiple data paths between the 202 and the texture memory 204, the number of pins of the processor can be greatly reduced, thereby reducing the printed circuit board cost.

Claims (1)

First storage means for storing texture elements (hereinafter, referred to as texels) necessary for obtaining pixel values constituting a screen; Temporary storage means for temporarily storing a predetermined number of texels; First address generating means for generating a transverse address of the texel necessary to obtain the pixel value; Second address generating means for generating a longitudinal address of the texel necessary to obtain the pixel value; Input the predetermined number of texels from the first storage means and output them to the temporary storage means, calculate a texel address corresponding to the mapping method and the transverse address and the longitudinal address, and the texel corresponding to the texel address is A control means for checking whether the temporary storage means exists and outputting the corresponding texel input from the temporary storage means or the corresponding texel input from the first storage means as a first color value according to the checked result. ; Shading means for itself generating a second color value necessary for shading for adding visual information to the image; Mixing means for inputting and mixing the first color value and the second color value, and outputting the mixed result as a third color value finally needed to obtain the pixel value; Second storage means for temporarily storing the third color value input from the mixing means as image data; And image display means for inputting the image data stored in the second storage means and displaying the image data on the screen.