Computer graphics processing and selective visual display system – Computer graphics processing – Three-dimension
Reexamination Certificate
1999-05-03
2002-02-12
Vo, Cliff N. (Department: 2671)
Computer graphics processing and selective visual display system
Computer graphics processing
Three-dimension
C345S420000
Reexamination Certificate
active
06346939
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to the field of computer graphics, and more particularly to a method for ordering objects that occupy a co-incident surface.
2. Description of the Related Art
Modem computer systems are widely used to model and render three-dimensional (3D) objects on two-dimensional (2D) display screens. For example, a variety of jobs that were usually drawn by hand (e.g., movie animation, architectural drawings, engineering drafting) are now performed with sophisticated computer graphics applications. These modern 3D computer graphics applications are able to render 3D images of not only a single fixed image, but also render views of the image dynamically as viewpoint and direction are changed. The term “viewpoint” refers to the point at which the viewer is located and is sometimes referred to as camera position, eye position, center of projection, or the like. The term “view direction” refers to the direction of view from the viewpoint and is commonly referred to as camera direction, eye direction, direction of projection, or the like. As the viewpoint and view direction change from one frame to the next, the 3D images are rendered dynamically to reflect the scene relative to the new viewpoint and the direction.
The changes in viewpoint and direction may occur in several forms in 3D viewing. For example, an object in a 3D space may be viewed from any spatial position: front, back, above, below, or side. Furthermore, an object such as a building, sphere, or the like may be viewed from the inside or outside of the object. These changes in viewpoint and direction directly affect the way an object is to be described and rendered.
For example, in 3D computer graphics, two or more objects are often drawn on a same plane in space when they essentially lie on the same plane.
FIG. 1A
illustrates a perspective view of a table
102
with a plurality of cards
106
. The cards
106
are placed on a top surface
104
of the table
102
. When viewed from a view direction
108
above the table
102
, the cards
106
appear on top of the surface
104
of the table
102
. However, when viewed from a view direction
110
below the table
102
, the cards
106
are not visible. Similarly,
FIG. 1B
shows a road surface
120
having a plurality of median stripes
122
painted thereon. The stripes
122
are visible when viewed from a view direction
124
above the road surface
120
. On the other hand, they are not visible when viewed from a view direction
126
below the road surface.
In rendering the table
102
with the cards
106
or the road surface
120
with the stripes
122
, conventional techniques have typically modeled the cards
106
and the surface
104
of the table
102
or the stripes
122
and the road surface
120
as co-planar polygons. Rendering these objects as being co-planar polygons, however, produces jitter about the co-planar regions. Since the layers are co-planar, the question “which layer is in front?” has no single answer. So algorithms that advance pixel-by-pixel across the screen, rendering the front-most surface at each pixel, are subject to error. Each pixel's query may return any answer, and adjacent pixels may display colors from different layers in the surface. This visual discontinuity across the surface is commonly referred to as jitter, z-tearing, or z-buffer tearing. If no enhancements are made from the basic algorithm, then techniques such as ray-tracing and z-buffer rendering will produce errors. The set of techniques that produce these errors, however, is not restricted to those of ray-tracing and z-buffer rendering. In some instances, the mathematical modeling of the co-planar polygons leads to z-buffer tearing or other degrading artifacts.
Additionally, the conventional methods have suffered from discontinuous or choppy images. For example, in one frame, the card
106
may appear to be over the top surface
104
of the table
102
. In another frame, the card
106
may actually appear to be under the table
102
. The inconsistent rendering of the objects thus degrades graphics image quality and produces unrealistic images.
To address such problems, one method has rendered two or more co-planar objects by rendering one object slightly above or below the other object. While this solution improves the jitter effects, it is highly sensitive to the separation distance between the two objects. For instance, if the distance between the two objects is too far, one object may appear to be floating over or hanging below the other due to the separation between them. On the other hand, if the distance were too small, the previously described jitter effect may be produced. In either case, the resulting image would not accurately convey the desired effect of one object resting upon another.
Another method has implemented a fixed stacking order technique. This method specifies the order of a number of layers (e.g., co-planar objects) and then renders the layers in the specified order.
FIG. 2A
shows a node tree
200
depicting a fixed stacking order
202
. The node tree
200
comprises a layer group node
204
, which branches out to a plurality of nodes
206
,
208
,
210
, and
212
. The nodes
206
,
208
,
210
, and
212
represent layer “A,” layer “B,” layer “C,” and layer “D,” respectively. The node tree specifies the stacking order
202
from left to right, i.e., from layer “A” to layer “D.” Accordingly, the node tree
200
is traversed in the order of nodes
206
,
208
,
210
, and
212
to render the layers from the lowest to the highest layer. Each layer is allowed to draw over any previous layer at all points, without comparing the relative distance from the eye. The result is layers that appear to lie correctly atop each other.
FIG. 2B
illustrates an orthographic view of the stacking order
202
of the layers “A,” “B,” “C,” and “D.” As shown, the layers are specified and rendered from back to front, i.e., from lowest to the highest. Thus, the lowest layer “A” is rendered first, then the second layer “B” is rendered on top of layer “A.” Next, the third layer “C” is drawn over layer “B” and finally, the highest layer “D” is rendered over the layer “C.”Unfortunately however, the fixed order of the conventional method does not allow rendering of a surface having two distinct sides from two distinct viewpoints. This is because the stacking order is fixed regardless of changes in view parameters such as viewpoint, view direction, or the like. For example, the stacking order
202
of
FIGS. 2A and 2B
only provides back-to-front drawing order of a surface and thus does not provide for jitter-free rendering of the other side of the surface as viewed from the opposite side. Hence, the problem of rendering co-planar or co-incident surface layers from other viewpoints is not adequately addressed by the conventional method.
In view of the foregoing, what is needed is a method for specifying and changing the ordering of co-planar or co-incident surface layers in response to changes in viewpoint or view direction without undesirable jitter or aliases.
SUMMARY OF THE INVENTION
The present invention fills these needs by providing method and system for determining view dependent ordering of geometry layers that occupy a co-incident surface. It should be appreciated that the present invention can be implemented in numerous ways, including as a process, an apparatus, a system, a device, a method, or a computer readable medium. Several inventive embodiments of the present invention are described below.
In one embodiment, the present invention provides a method for determining view dependent ordering of geometry layers that occupy a co-incident surface. The co-incident surface has a first side and a second side, and is capable of being viewed on either side. In this method, a first order of the geometry layers is defined as viewed on the first side of the co-incident surface. In response to a specified view, a rendering order of the geometry layers is determined from the first order. The rend
LandOfFree
View dependent layer ordering method and system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with View dependent layer ordering method and system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and View dependent layer ordering method and system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2941474