Computer graphics processing and selective visual display system – Computer graphics processing – Attributes
Reexamination Certificate
2001-09-18
2004-09-14
Bella, Matthew C. (Department: 2676)
Computer graphics processing and selective visual display system
Computer graphics processing
Attributes
C345S426000, C345S622000
Reexamination Certificate
active
06791563
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to computer graphics systems, and more specifically to a method for global rendering of multidimensional textures.
2. Related Art
Computer rendering is the art of representing a three dimensional (3D) object or scene in the two dimensional (2D) space of a computer screen or other flat medium. Vast increases in computing power have resulted in increasingly complex and sophisticated computer applications. Such applications often call for equally sophisticated and complex computer rendering. Typically, 3D objects are rendered, or modeled, as a collection of joined polygons (typically triangles) defined by vertex positions and attributes. The end result of the rendering process is a collection of pixels for each surface of the projected polygon.
In early rendering processes, designers would take a picture of an object, such as a tree or a person, and simply map that image to a vertical plane oriented towards the viewpoint in their scene. In this fashion, one could place various objects into the scene and have them appear natural. This technique is analogous to placing a number of billboards into the scene, each oriented toward the viewer. This worked well for static scenes but, for the dynamic effects required by animation, and for more complex lighting schemes, it was clear that the objects are really two dimensional billboards.
Traditionally, designers wishing to render a 3D object, such as a person or a tree, would take a series of photographs of that object, for example, around an entire 360 degree panorama with each photograph offset by one or two degrees from the previous photograph. This plurality of images was then combined into a single file. Such a file is often referred to as a multidimensional texture (MDT). When the rendering operation was performed, the software would place an empty plane representing the object in the scene to be rendered, and would select, from the plurality of images in the file, the image corresponding to the current viewpoint, i.e. the image that was facing the viewer. In this fashion, if the viewer is looking at the front of the object, the software will pick the image representing that side. If the viewer were looking at the back of the object, the software would pick the image representing the back of the object. It will be appreciated, however, that this image is still just a two dimensional plane in the scene, i.e., it is still just like a billboard.
Because images are still represented by a plane or a billboard, it is difficult to perform complex lighting and shading operations in the scene. This is particularly true when the scene contains a plurality of light sources, or when the light source is perpendicular to the 2D billboard representing the object. These sorts of complex rendering operations are referred to herein as “global rendering.” It will be appreciated that such global rendering has the potential to enhance realism by modeling real world visual effects such as shadows, surface shading, and illumination from various different sources of light.
The disadvantages of the billboard technique are illustrated in
FIGS. 1A and 1B
. Each illustration is an actual rendering using a conventional rendering method, and each illustration represents the same view of a multi-dimensional texture object
104
placed in a room with two mirrored walls
112
and
114
. The large object in the middle of the view is the object as seen from the viewpoint location. The other three objects
106
,
108
and
110
in each view are reflections of the object
104
in the mirrored walls.
FIG. 1A
shows the single, static rectangle used by the conventional method to render MDTs. For each MDT object, a vertical rectangle
104
is placed in the scene
102
and oriented so that it is perpendicular to the view direction. A single texture image from the MDT file is extracted and mapped to this rectangle. A shadow
116
is cast from a light source that is perpendicular to the view direction.
FIG. 1B
shows a single texture from the MDT file mapped to the rectangle in FIG.
1
A. The foreground object
118
displays the correct side view of the texture object. However, there are several problems inherent with this method when global light effects are considered. First, the reflections of the foreground object
120
,
122
and
124
in the mirrored walls incorrectly display the same side view of the texture as the foreground object. Second, the reflected images of the object (especially reflection
124
) are distorted because the reflected planes are not perpendicular to the view direction. Third, the shadow
126
cast by the foreground object is extremely thin. This is because the direction of the light source is almost parallel to the plane of the foreground object
118
. The limitations of the established method result in incorrect and distorted views of the object from the other view locations.
It would be desirable, therefore, to provide a method of rendering that allows more realistic global illumination and rendering of 3D scenes.
SUMMARY OF THE INVENTION
In an exemplary embodiment of the present invention, a system, method and computer program product for global rendering of multidimensional textures (MDTs) using bounded volumes are disclosed. By using bounded volumes, objects to be rendered receive more realistic representation in a scene.
It is an advantage of the present invention that an image in a computer can be rendered realistically from any view simultaneously. By using bounding geometry combined with multi-dimensional texture files, the present invention eliminates the problems of incorrect shadows, distorted views and incorrect lighting effects that existed with previous techniques of rendering that only rendered one view at a time.
In an exemplary embodiment, the present invention can be a method in a computer system for global rendering of a multidimensional texture map including the steps of placing an object representing a multidimensional texture map into a database, replacing the object with a bounding geometry, sampling the bounding geometry with a sample of interest having an origin and a direction, extracting pixel information used for global rendering of the object from the sample of the bounding geometry, and displaying the object using the extracted pixel information. The bounding geometry can be a volume that completely encloses the object, and can be a cylinder with height equal to the object, and diameter equal to the width of the object, and the cylinder can be capped by a hemisphere of the same diameter. The sample of interest can be determined by an external rendering engine, and the origin can be for example, a camera, a light source or a point on a reflecting surface.
The pixel information extraction step can further include the steps of computing a direction from the center of the bounding geometry to the origin of the sample of interest, identifying a vertical plane that contains the center point of the bounding geometry and that is perpendicular to the sample of interest, projecting the sample onto the vertical plane, computing a coordinate of the projected sample in the vertical plane, extracting an image from a multi-dimensional texture where the image faces the computed direction, extracting a pixel from the image that corresponds to the coordinate, and extracting texture information about the pixel. The texture information can be the color, transparency, Z-depth, or surface normal of the extracted pixel.
Computing a coordinate of the projected sample in the vertical plane can include setting a coordinate origin at the lower left corner of the vertical plane as viewed from the origin of the sample of interest, computing a U coordinate of the projected sample as the horizontal distance from the coordinate origin to the projected sample, and computing a V coordinate of the projected sample as the vertical the coordinate origin to the projected sample.
In an alternative exemplary embodiment, the present invention can be a system th
Bragg Dennis
Segal Peter
Bella Matthew C.
Bentley Systems Incorporated
Caschera Antonio
Kaminski Jeffri A.
Swindell Caroline
LandOfFree
System, method and computer program product for global... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System, method and computer program product for global..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System, method and computer program product for global... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3205524