WO2001056268A1

WO2001056268A1 - Method for rapid smoothing of object edges in computer graphics

Info

Publication number: WO2001056268A1
Application number: PCT/US2001/002794
Authority: WO
Inventors: Amon Tavor
Original assignee: Channel Storm Ltd.; Friedman, Mark, M.
Priority date: 2000-01-27
Filing date: 2001-01-29
Publication date: 2001-08-02
Also published as: AU2001231212A1

Abstract

A method for rapidly performing an edge anti-aliasing procedure on a computer graphic image, by simultaneously performing a texture filtering procedure, thereby enabling both texture anti-aliasing and edge anti-aliasing to be performed. According to a first embodiment, the value of the alpha component is calculated for each pixel individually in order to performed the edge anti-aliasing method (Fig.1). According to preferred embodiments, an extra row or column of pixels is drawn on each side of the image with their alpha component set to total transparency, but with the same color value as the nearest pixels of the original image (Fig.5), to form an enhanced image. If the original pixels of the image had no alpha values, then these alpha values are set to full opacity. When the enhanced image is transformed, for example by being rotated and slanted, texture filtering is then performed, for example by linear interpolation (Fig.6).

Description

METHOD FOR RAPID SMOOTHING OF OBJECT EDGES IN COMPUTER GRAPHICS

FIELD AND BACKGROUND OF THE INVENTION The present invention relates to a method for rapid smoothing of object edges in computer graphics, and in particular, to such a method in \\ hich anti-aliasing procedures are performed with texture filtering procedures.

Standard computer display systems are able to display graphic images as a combination of straight rows and columns of rectangular display segments, or picture elements, known as pixels. Each combination of pixels results in an image by setting pixels within the combination to different colors, for color images. The graphic image can then be stored in computer memon as a list of pixel values, such that producing the image on the computer

involves copying the pixel values to the appropriate pixel locations on the computer display.

Image processing involves various steps for manipulating and altering the graphic image, particularly for the processing of color images. For example, the graphic image could be scaled or rotated. However, there may not be an exact correspondence between each pixel in the graphic image as stored in memory, and the pixels of the computer display system, or "screen pixels". Therefore, when mapping each image pixel to a screen pixel, the result is sometimes rounded in order to choose an appropriate image pixel, which is known as the "nearest neighbor" method. This method is fast, but often results in unwanted visual effects known as "aliasing". Aliasing effects can be divided into two main categories: texture aliasing and edge aliasing. Texture aliasing occurs when textured images are transformed and displayed on the computer display screen in the "nearest neighbor" method. Some of the image information may be lost during the process of transformation, and unwanted patterns may appear on the image. For example, if an image contains a straight line which is only one pixel wide, the application of the "nearest neighbor" method may result in the line disappearing altogether from the image as displayed on the screen, due to texture aliasing.

Such texture aliasing is demonstrated in background art Figures 1 and 2. Background art Figure 1 shows a sample correspondence between image pixels and screen pixels after transformation has been performed, but without correction. As shown, there are twenty image pixels, of which every fifth pixel is black. These image pixels are mapped to the screen pixels with the nearest neighbor method. Thus, when displayed on the computer display screen, the image has been reduced by twenty-five percent to cover an area of only fifteen pixels, resulting in the corresponding loss of twenty- five percent of the black pixels.

Background art Figure 2 shows the same transformation between image pixels and screen pixels, but with the addition of a corrective procedure against texture aliasing. Texture aliasing can be overcome by one of several different such corrective procedures, generally referred to as "texture filtering". One common corrective method, the effect of which is shown in background art Figure 2, is known as linear interpolation. Linear interpolation involves the calculation of screen pixel values by interpolating values from two (or more) image pixels when there is no exact match between screen pixels and image pixels. As shown in background art Figure 2, the resultant image after linear interpolation may be somewhat blurred, as certain screen pixel values represent an average value rather than an absolute value. These average values are given in the key: if the image pixel values are one hundred percent, then the corresponding screen values range from twenty-five percent to seventy-five percent. However, at least the loss of image information shown in background art Figure 1 has been avoided. Furthermore, texture aliasing has the advantage of being performed relatively quickly, since the required calculations are of low complexity, and the resulting pixel values are only calculated from values of adjacent pixels in the same image.

Background art Figures 3 A and 3B show the effects of edge aliasing, which is another type of aliasing effect. Edge aliasing commonly appears when images are transformed by being rotated or slanted before being displayed on the computer display screen. The edges of such images appear "jagged" or uneven because the angle of the image edge is not aligned with the grid of the rows and columns of the screen pixels. Background art Figures 3 A and 3B show a section of such a rotated image. Background art Figure 3 A shows the rotated image without any color values, while background art Figure 3B shows the same image with black pixel values, as it would appear on the computer screen. The jagged edges of the rotated image can clearly be seen.

Edge aliasing may be overcome through an "anti-aliasing" method, in which the level of opacity of each edge pixel of the image is calculated according to the theoretical area of the pixel covered by the image. Pixels which are entirely covered become opaque, while pixels which are only partially covered become blended with the colors of the background of the image, such that the level of the color for each pixel is in linear proportion to the area of that pixel which lies within the image. For example, if the image is composed of black pixels and the background is composed of white pixels, then the value of pixels which are seventy-five percent inside the image becomes seventy-five percent black; the value of pixels which are fifty percent inside the image becomes fifty percent black; and so forth. Background art Figure 4 shows the result of performing such an anti-aliasing method on the image of Figure 3B. As shown in the accompanying key, the value of the pixels ranges from one hundred percent, for pixels which lie completely within the image, to twenty percent, for pixels which are only fractionally within the image. The disadvantage of edge anti-aliasing techniques is that they are complex to calculate, involving both background pixel values and image pixel values, unlike texture filtering techniques. Furthermore, edge anti-aliasing methods may vary with the shape of the image, such that calculating edge pixel opacity values for a rectangular image is different than calculating such values for a round image, for example. Therefore, there are no rapid, automatic methods for performing edge anti-aliasing on computer graphics. One commonly used method for overcoming edge anti-aliasing on computer graphic images is to drawing each image to an off-screen buffer in a large scale, for example four or sixteen times larger than the required image. The resulting image is then redrawn from the off-screen buffer to the computer display screen. In order to smooth the edge pixels, texture filtering can be used to interpolate adjacent pixels, since as previously mentioned, texture filtering methods may be performed relatively rapidly. However, the entire method for overcoming edge aliasing considerably increases the time required to generate the computer graphic image. Furthermore, this problem cannot be overcome by graphic acceleration hardware, which is currently used in many computer systems in order to increase the speed of drawing graphic images onto the computer display screen. Such hardware can be used for rapid texture filtering, for example, since the calculations involved are relatively simple. However, for the previously described reasons, such hardware cannot currently be used for accelerating the performance of edge anti-aliasing methods.

The result of this limitation is apparent in many different applications. For example, in computer games with three-dimensional graphics, graphic images are transformed in order to draw textured objects in a three-dimensional environment. Such applications often use the texture filtering offered by graphic acceleration hardware in order to perform texture anti-aliasing methods, but cannot perform edge anti-aliasing methods since the graphics must be rapidly drawn. Thus, edges of objects often appear jagged or uneven when drawn on the computer screen. A useful solution to this problem would provide a rapid method for performing edge anti-aliasing procedures when drawing transformed graphic images to a computer display screen. Such a method would also preferably be suitable for use with existing graphic acceleration hardware, thereby enabling edge anti-aliasing techniques to be performed for a wide spectrum of applications which cannot currently take advantage of such techniques, such as the previously described computer game applications. Unfortunately, such a method does not currently exist.

There is thus a need for. and it would be useful to have, a method for rapidly performing edge anti-aliasing procedures on graphic images, which could also be used with existing graphic acceleration hardware, thereby enabling edge anti-aliasing methods to be performed with a wide variety of computer graphic applications.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing and other objects, aspects and advantages will be better understood from the following detailed description of a preferred embodiment of the invention with reference to the drawings, wherein:

FIG. 1 is a schematic block diagram of a background art image, showing a sample correspondence between image pixels and screen pixels after transformation has been performed, but without correction;

FIG. 2 is a schematic block diagram of the background art image of Figure 1 , showing the same transformation between image pixels and screen pixels, but with the addition of a corrective procedure against texture aliasing; FIGS. 3 A and 3B show the effects of edge aliasing with background art schematic block diagrams of a rotated image;

FIG. 4 is a background art of a schematic block diagram, showing the result of performing an anti-aliasing method on the image of Figure 3B; FIGS. 5 and 6 show an exemplary original image before (Figure 5) and after (Figure 6) the performance of the first portion of a preferred method according to the present invention; and

FIGS. 7 A and 7B show an exemplary original image after the performance of the preferred method according to the present invention, without (Figure 7A) and with (Figure 7B) color and alpha component values.

SUMMARY OF THE INVENTION

The present invention is of a method for rapidly performing an edge anti-aliasing procedure on a computer graphic image, by simultaneously performing a texture filtering procedure, thereby enabling both texture anti-aliasing and edge anti-aliasing to be performed. The method of the present invention uses the alpha component of color values for the pixels in order to adjust the degree of transparency of edge pixels, thereby rendering edges which are smooth in appearance.

According to a first embodiment of the present invention, the value of the alpha component is calculated for each pixel individually in order to perform the edge anti-aliasing method. The calculation is performed by determining the value for the alpha component according to the theoretical area of the screen pixel which is covered by the visible shape of the image, such that when a greater proportion of the screen pixel is covered by the image, the value for the alpha component is correspondingly increased for greater opacity of the pixel. However, this embodiment of the method of the present invention still requires the calculation of the relative fraction of each screen pixel which is covered by the image, and as such still requires extensive computational resources and time to perform.

According to preferred embodiments of the method of the present invention, an extra row or column of pixels is added to each side of the image in memory prior to the image being drawn on the screen. The extra row or column of pixels has their alpha component value set to total transparency, but with the same color value as the nearest pixels of the original image, to form an enhanced image. If the original pixels of the image had no alpha values, then these alpha values are set to full opacity. When the enhanced image is transformed, for example by being rotated or slanted, texture filtering is then performed, for example by linear interpolation. When texture filtering is performed, the alpha component values of those pixels which were previously on the edge are automatically calculated by interpolation of the alpha component of these pixels with the alpha component of the enhanced pixels, in proportion to the area of these pixels which are covered by each calculated screen pixel. Thus, the edges of the graphic image appear smooth, as though an edge anti-aliasing method had been performed, but with the speed and ease of calculation of a texture anti-aliasing method.

According to the present invention, there is provided a method for performing edge anti-aliasing on an image to form a smooth image, the image comprising a plurality of pixels, the steps of the method being performed by a data processor, the method comprising the steps of: (a) assigning an alpha component value to at least a portion of the plurality of pixels; (b) adjusting the alpha component value for at least one pixel having the alpha component value according to the alpha component value of at least one additional pixel to form an anti-aliased pixel; and (c) displaying the plurality of pixels, including the anti-aliased pixel, to form the smooth image.

According to another embodiment of the present invention, there is provided a method for performing edge anti-aliasing on an image to form a smooth image, the image comprising a plurality of pixels, each pixel having an alpha component value, the steps of the method being performed by a data processor, the method comprising the steps of: (a) adjusting the alpha component value of at least one pixel according to the alpha component value of at least one adjusting pixel to form an anti-aliased pixel; and (b) displaying the plurality of pixels, including the anti-aliased pixel, to form the smooth image.

According to yet another embodiment of the present invention, there is provided a method for performing edge anti-aliasing on an image to form a smooth image, the image comprising a plurality of pixels, each pixel having an alpha component value, the steps of the method being performed by a data processor, the method comprising the steps of: (a) determining a plurality of edge pixels of the image; and (b) interpolating between the alpha component value of the edge pixel and an alpha component value for an adjusting pixel to form an edge anti-aliased pixel.

According to still another embodiment of the present invention, there is provided a method for performing edge anti-aliasing on an image to form a smooth image, the image comprising a plurality of pixels, the steps of the method being performed by a data processor, the method comprising the steps of: (a) determining a plurality of edge pixels of the image; (b) determining a plurality of adjusting pixels for the image; (c) assigning an alpha component value to each of the plurality of edge pixels and the plurality of adjusting pixels; and (d) interpolating between the alpha component value of the edge pixel and the alpha component value for the adjusting pixel to form an edge anti-aliased pixel.

Hereinafter, the term "computer" includes, but is not limited to, personal computers (PC) having an operating system such as DOS, Windows™, OS/2™ or Linux; Macintosh™ computers; computers having JAVA™-OS as the operating system; and graphical workstations such as the computers of Sun Microsystems™ and Silicon Graphics™, and other computers having some version of the UNIX operating system such as AIX™ or SOLARIS™ of Sun Microsystems™; or any other known and available operating system. Hereinafter, the term "Windows™" includes but is not limited to Windows95™, Windows 3.x™ in which "x" is an integer such as "1", Windows NT™, Windows98™. Window s CE™ and any upgraded versions of these operating systems by Microsoft Corp. (USA).

The method of the present invention could be described as a series of steps performed by a data processor, and as such could optionally be implemented as software, hardware or firmware, or a combination thereof. For the present invention, a software application could be written in substantially any suitable programming language, which could easily be selected by one of ordinary skill in the art. The programming language chosen should be compatible with the computer hardware and operating system according to which the software application is executed. Examples of suitable programming languages include, but are not limited to, C, C++ and Java.

DETAILED DESCRIPTION OF THE INVENTION

The present invention is of a method for rapidly performing an edge anti-aliasing procedure on a computer graphic image, by simultaneously performing a texture filtering procedure, thereby enabling both texture anti-aliasing and edge anti-aliasing to be performed. Specifically, the edge anti-aliasing method involves the use of the alpha component of the pixel color values. The alpha component of the pixel color values is automatically inteφolated for neighboring pixels during the performance of texture filtering. This interpolation is used for the method of the present invention to also perform edge anti-aliasing. Optionally, edge anti-aliasing is performed without texture filtering, by calculating the alpha component value for each pixel individually, although this method is less preferred, as described in greater detail below.

The alpha component is an additional component of the color values for color image pixels. For example, each color image pixel optionally has three color components in a three color component system, such as the RGB (red, green, blue) system. The alpha component then forms the fourth component, and is used to determine the degree of transparency or opacity of the pixel. For example, the alpha component may be used to cause certain areas in the image to appear transparent or semi-transparent when images are drawn on top of each other, or on top of a background color. According to a first embodiment of the present invention, the value of the alpha component is calculated for each pixel individually in order to perform the edge anti-aliasing method. The calculation is performed by determining the value for the alpha component according to the theoretical area of the screen pixel which is covered by the visible shape of the image, such that when a greater proportion of the screen pixel is covered by the image, the value for the alpha component is correspondingly increased for greater opacity of the pixel. Therefore, when the first image is drawn on top of a second image or on top of a colored background, the color values for this first image are blended with those of the second image or of the colored background according to the value for the alpha component of each pixel of the first image. Such blending results in image edges which are smooth in appearance. However, this embodiment of the method of the present invention still requires the calculation of the relative fraction of each screen pixel which is covered by the image, and as such still requires extensive computational resources and time to perform. Therefore this embodiment of the method is less preferred.

According to preferred embodiments of the method of the present invention, an extra row or column of pixels is added to each side of the image when stored in memory, which feature the alpha component value set to total transparency, but with the same color value as the nearest pixels of the original image, to form an enhanced image. If the original pixels of the image had no alpha values, then these alpha values are set to full opacity. When the enhanced image is transformed, for example by being rotated or slanted, texture filtering is then performed, for example by linear interpolation. When texture filtering is performed, the alpha component values of those pixels which were previously on the edge are automatically calculated by interpolation of the alpha component of these pixels with the alpha component of the enhanced pixels, in proportion to the area of these pixels which are covered by each calculated screen pixel. Thus, the edges of the graphic image appear smooth, as though an edge anti-aliasing method had been performed, but with the speed and ease of calculation of a texture anti-aliasing method.

This preferred embodiment of the method according to the present invention is operative because of the effect of texture filtering on images composed of pixels having an alpha component, which may be black and white (grayscale) images or color images. Texturing filtering methods, such as the linear interpolation method, involve the calculation of the color components of each displayed pixel from the inteφolation of the component values from the image pixels which cover the screen pixel. If the image pixels contain an alpha component, the value of the alpha component for the displayed pixel is also calculated. When the filtered pixel is drawn to the computer screen over another image or over a colored background, the color of the filtered pixel is blended with the background color or color of the second image according to the calculated value of the alpha component for the filtered pixel.

Therefore, according to the present invention, the same method for texture filtering performs the texture anti-aliasing procedure and the edge anti-aliasing procedure, by calculating the level of opacity of each pixel by inteφolation. Optionally, such a method according to the present invention is performed with graphic acceleration hardware, thereby enabling the method to be rapidly performed, such that a substantial increase in the time and computational resources required to draw the graphic image to the computer screen is not required.

Examples of suitable methods of texture filtering which are operative with the method of the present invention include, but are not limited to, linear inteφolation and bicubic inteφolation, and "shifting and blending". Linear inteφolation is performed by measuring the distance between the image pixels and the screen pixel. If the screen pixel fits exactly on the image pixel, such that there is zero distance between the screen pixel and the image pixel, then the image pixel is drawn "as is". If the screen pixel is equidistant between two image pixels, then the image pixels are blended equally. If the screen pixel is closer to one image pixel than to the second image pixel, the image pixel which is closer has greater weight in the blending calculation, preferably in inverse proportion to the distance from the screen pixel to the image pixel.

Bicubic inteφolation is another common method for texture filtering. This method is similar to the method of linear inteφolation, except that a bicubic function is used to blend the pixel values, rather than a linear function. The results are more accurate, but the method is computationally much more intensive than linear inteφolation, and is therefore only used in applications where quality is critical. Another exemplary method for texture filtering is "shifting and blending", in which the image is drawn several times, each time in a slightly different position, and all of the drawn images are blended together. This method is suitable for both texture anti-aliasing and edge anti-aliasing, but is very slow since the entire image must be redrawn many times in order to achieve reasonable results.

The principles and operation of a method according to the present invention may be better understood with reference to the drawings and the accompanying description, it being understood that these drawings are given for illustrative puφoses only and are not meant to be limiting. Referring now to the drawings. Figures 5 and 6 show a sample original image before (Figure 5) and after (Figure 6) the performance of the first portion of a preferred method according to the present invention, which is actually preprocessing for the texture filtering procedure. In the preferred embodiment of the method of the present invention, the original image of Figure 5 is enhanced by the addition of an extra row or column of pixels to each side of the image. The color value of each extra pixel is copied from the nearest neighboring pixel of the original image. The value of the alpha component for each extra pixel is set to total transparency, while the alpha component value for the pixels of the original image is preferably set to full opacity.

Alternatively, if the alpha component value of one or more pixels of the original image had been previously set to a value less than full opacity, then optionally the one or more pixels of the original image retain that previously set alpha value. Figure 5 shows the sample original image before the performance of the method according to the present invention. This original image features black pixels, displayed on white background pixels. In order to perform the method of the present invention, an extra row or column of pixels is added to each side of the image, as previously described. Figures 6A and 6B show the sample image of Figure 5 after the performance of the method according to the present invention to form an enhanced image. Figure 6A shows the pixel color values, while Figure 6B shows the pixel alpha values. Therefore, the color values of the extra pixels are identical to those of the nearest neighboring original pixels, while the alpha values of the extra pixels are set to total transparency. If the enhanced image is directly drawn to the computer screen, the drawing dimensions of the enhanced image are increased in proportion to the additional pixels, in order to maintain the visible scale of the image. However, since the additional pixels are fully transparent, they do not override the background color. Thus, if the enhanced image is not transformed, it would not appear differently than the original sample image when drawn to the computer screen.

However, when the enhanced image is transformed, for example by being rotated or slanted, then texturing filtering is performed, which is the second portion of the preferred method according to the present invention. The performance of texture filtering causes the alpha component values of the original edge pixels to be automatically calculated by inteφolation of the alpha component values for these pixels with the alpha component values for the additional pixels. Such inteφolation causes the alpha component values for the edge pixels of the final image to be determined according to the proportion of the area of the original edge pixels which is covered by each calculated screen pixel. Thus, the resultant edges of the final image appear smooth, as if a separate edge anti-aliasing method had been performed.

The effect of the method according to the present invention is illustrated with exemplary Figures 7 A and 7B. Exemplary Figure 7A shows a sample image without any color or alpha component values, superimposed on a grid. The original image, a rectangle, is outlined with a solid black line, while the enhanced image is outlined with a gray line. The extra pixels are therefore those pixels located between the inner rectangle (black line) and the outer rectangle (gray line).

Figure 7B shows the sample image of Figure 7 A, with the combined color and alpha component values after the performance of the preferred method of the present invention. These combined values are illustrated schematically, according to the associated key (shown in Figure 4), for the puφoses of illustration only and without any intention of being limiting. As described above, the preferred method involves the performance of texture filtering on the enhanced image. The inteφolated pixels form the new edge pixels of the filtered image, as shown. The appearance of the pixels of the final image is determined according to a combination of the color and alpha component values, particularly by the degree of transparency which is determined by the alpha component values. Thus, the final image would appear to have smooth edges.

According to a particularly preferred embodiment of the present invention, a hardware acceleration device is used for image processing with the preferred method of the present invention. Once the enhanced image had been performed, with the additional pixels, such an acceleration device is optionally and preferably used to perform the texture filtering portion of the method, for example by being used to perform linear inteφolation. Examples of suitable graphic hardware acceleration devices for use with the present invention may be obtained from such manufacturers as ATI Technologies, Diamond Multimedia, Creative Labs and 3Dfx. among others. Thus, the hardware acceleration device increases the speed and efficiency of the performance of the texture filtering, thereby enabling both texture filtering and edge anti-aliasing to be achieved in a single procedure, without significantly adding to the time or computational resources which are required for the texture filtering process alone.

It will be appreciated that the above descriptions are intended only to serve as examples, and that many other embodiments are possible within the spirit and the scope of the present invention.

Claims

WHAT IS CLAIMED IS:

1. A method for performing edge anti-aliasing on an image to form a smooth image, the image comprising a plurality of pixels, the steps of the method being performed by a data processor, the method comprising the steps of:

(a) assigning an alpha component value to at least a portion of the plurality of pixels;

(b) adjusting said alpha component value for at least one pixel having said alpha component value according to said alpha component value of at least one additional pixel to form an anti-aliased pixel; and

(c) displaying the plurality of pixels, including said anti-aliased pixel, to form the smooth image.

2. The method of claim 1, wherein step (b) is performed with a texture filtering method for inteφolating between said alpha component value of said at least one pixel and said alpha component value for said at least one additional pixel.

3. The method of claim 1, wherein the image features a plurality of edge pixels, wherein said plurality of edge pixels are assigned alpha component values in step (a) and wherein step (a) further comprises the steps of:

(i) adding an added pixel next to each edge pixel, each added pixel having an alpha component value; and (ii) adjusting said alpha component value of each edge pixel according to said alpha component value of said added pixel.

4. The method of claim 3, wherein step (ii) is performed with a texture filtering method for inteφolating between said alpha component value of said edge pixel and said alpha component value for said added pixel.

5. The method of claim 3, wherein step (i) further comprises the steps of:

(1) setting said alpha component value for said added pixel to full transparency: and

(2) setting said alpha component value for said plurality of edge pixels to at least partial opacity.

6. The method of claim 5, wherein step (2) is performed by setting said alpha component value for said plurality of edge pixels to full opacity.

7. The method of claim 4, wherein said texture filtering method comprises the steps of:

(1) redrawing the image with said added pixels at least once to form at least one redrawn image, such that said at least one redrawn image is shifted in location with regard to the image: and

(2) blending the image with said at least one redrawn image.

8. The method of claim 1, wherein step (b) further comprises the step of:

(i) providing a hardware accelerator device, such step (b) is performed by said hardware accelerator device.

9. The method of claim 1, wherein step (b) further comprises the steps of: (i) determining a plurality of edge pixels of the image;

(ii) determining a proportion of each edge pixel covered by a neighboring pixel; and (iii) determining an alpha component value for each edge pixel according to said proportion.

10. A method for performing edge anti-aliasing on an image to form a smooth image, the image comprising a plurality of pixels, each pixel having an alpha component value, the steps of the method being performed by a data processor, the method comprising the steps of:

(a) adjusting the alpha component value of at least one pixel according to the alpha component value of at least one adjusting pixel to form an anti-aliased pixel; and

(b) displaying the plurality of pixels, including said anti-aliased pixel, to form the smooth image.

11. The method of claim 10, wherein step (a) is performed with a texture filtering method for inteφolating between the alpha component value of said at least one pixel and the alpha component value for said adjusting pixel.

12. The method of claim 10, wherein the image features a plurality of edge pixels, and wherein step (a) further comprises the steps of:

13. The method of claim 12, wherein said alpha component value of each added pixel is set to full transparency.

14. The method of claim 12. wherein step (ii) is performed with a texture filtering method for interpolating between the alpha component value of said edge pixel and said alpha component value for said added pixel.

15. The method of claim 14, wherein said texture filtering method comprises the steps of:

(1) redrawing the image with said added pixels at least once to form at least one redrawn image, such that said at least one redrawn image is shifted in location with regard to the image; and

(2) blending the image with said at least one redrawn image.

16. The method of claim 14, wherein step (ii) further comprises the step of:

(1) providing a hardware accelerator device, such that said texture filtering method is performed by said hardware accelerator device.

17. The method of claim 10, wherein step (a) further comprises the steps of:

(i) determining a plurality of edge pixels of the image;

(ii) determining a proportion of each edge pixel covered by a neighboring pixel; and (iii) determining the alpha component value for each edge pixel according to said proportion.

18. A method for performing edge anti-aliasing on an image to form a smooth image, the image comprising a plurality of pixels, each pixel having an alpha component value, the steps of the method being performed by a data processor, the method comprising the steps of:

(a) determining a plurality of edge pixels of the image; and

(b) inteφolating between the alpha component value of said edge pixel and an alpha component value for an adjusting pixel to form an edge anti-aliased pixel.

19. The method of claim 18, wherein said adjusting pixel is formed by adding an additional pixel to the image next to each edge pixel, each additional pixel having an alpha component value.

20. The method of claim 19, wherein said alpha component value for each additional pixel is set to full transparency.

21. The method of claim 18, wherein step (b) is performed with a texture filtering method.

22. The method of claim 18, further comprising the step of:

(c) displaying the plurality of pixels, including said edge anti-aliased pixel, to form the smooth image.

23. A method for performing edge anti-aliasing on an image to form a smooth image, the image comprising a plurality of pixels, the steps of the method being performed by a data processor, the method comprising the steps of:

(a) determining a plurality of edge pixels of the image;

(b) determining a plurality of adjusting pixels for the image;

(c) assigning an alpha component value to each of said plurality of edge pixels and said plurality of adjusting pixels; and (d) inteφolating between said alpha component value of said edge pixel and said alpha component value for said adjusting pixel to form an edge anti-aliased pixel.

24. The method of claim 23, wherein said adjusting pixel is formed by adding an additional pixel to the image next to each edge pixel, each additional pixel having an alpha component value.

25. The method of claim 24, wherein said alpha component value for each additional pixel is set to full transparency, and wherein said alpha component value for each edge pixel is set to at least partial opacity.

26. The method of claim 25, wherein said alpha component value for each edge pixel is set to full opacity.

27. The method of claim 23, wherein step (d) is performed with a texture filtering method.

28. The method of claim 23, further comprising the step of:

(e) displaying the plurality of pixels, including said edge anti-aliased pixel, to form the smooth image.