Computer graphics processing and selective visual display system – Computer graphics processing – Three-dimension
Reexamination Certificate
1999-10-04
2002-05-07
Vo, Cliff N. (Department: 2671)
Computer graphics processing and selective visual display system
Computer graphics processing
Three-dimension
Reexamination Certificate
active
06384821
ABSTRACT:
DESCRIPTION
Background of the Invention
1. Field of the Invention
The present invention generally relates to graphics processing and display systems and, more particularly, to the creation and presentation of three-dimensional scenes of synthetic content stored on distributed network sources and accessed by computer network transmission. The invention further relates to methods of adaptively selecting an optimal delivery strategy for each of the clients based on available resources.
2. Background Description
Using three-dimensional graphics over networks has become an increasingly effective way to share information, visualize data, design components, and advertise products. As the number of computers in the consumer and commercial sectors with network access increases, the number of users accessing some form of three-dimensional graphics is expected to increase accordingly. For example, it has been estimated by W. Meloni in “The Web Looks Toward 3D”,
Computer Graphics World,
21(12), December 1998, pp. 20 et seq., that by the end of year 2001, 152.1 million personal computers (PCs) worldwide will have an Internet connection. Out of this number, approximately 52.3 million users will frequently access three-dimensional images while on the World Wide Web (WWW or the Web). This number compares to only 10 million users accessing three-dimensional Web images in 1997 out of a total of 79 million Internet users. However, the use of three-dimensional graphics over networks is not limited to consumer applications. In 1997, roughly 59% of all U.S. companies had intranet connections. By 2001 this figure is expected to jump to 80%. This transition includes three-dimensional collaboration tools for design and visualization. For instance, within the computer-aided design (CAD) community there is significant interest in applications which permit sharing on a global basis of three-dimensional models among designers, engineers, suppliers and other interested parties across a network. The capability to perform “visual collaborations” offers the promise to reduce costs and to shorten development times. Other corporate interests target the use of three-dimensional solutions to visualize data such as financial fluctuations, client accounts, and resource allocations.
As generally shown in
FIG. 1
, three-dimensional models and their representations are typically stored on centralized servers
100
and are accessed by clients
101
over communication networks
102
. Several data-transfer technologies have been developed over the past few years to visualize three-dimensional models over networks.
At one end of the spectrum are the so-called client-side rendering methods in which the model is downloaded to the client which is entirely responsible for its rendering.
FIG. 2
shows a diagram of a typical client-side rendering architecture. Upon input from a user or another application
201
, the client
202
requests, via network
203
as client feedback
204
, a model from the server
205
. The geometry server
210
within server
205
contains the 3d geometry
211
and the scene parameters
212
. In response to client feedback
204
, the server
205
retrieves the model from storage
206
and delivers the 3d geometry
213
to the client
202
over the network
203
. Once the model has been received by the client, the client
3
d browser
208
renders it in client rendering engine
207
and displays it on the display
209
. Additional client feedback may follow as the user interacts with the model displayed and more information about the model is downloaded. Such methods typically require a considerable amount of time to download and display on the client an initial meaningful representation of a complex three-dimensional model. These methods also require the existence of three-dimensional graphics capabilities on the client machines.
Alternatives to en masse downloading of a model without prior processing include storage and transmission of compressed models, as reported by G. Taubin and J. Rossignac in “Geometry Compression Through Topological Surgery”,
ACM Transactions on Graphics
, April 1998, pp. 84-115, streaming and progressive delivery of the component geometry, as reported by G. Taubin et al. in “Progressive Forest Split Compression”,
ACM Proc. Siggraph '
98, July 1998, pp. 123—132, H. Hoppe in “Progressive Meshes”,
ACM Proc. Siggraph '
98, August 1996, pp. 99-108, and M. Garland and P. Heckbert in “Surface Simplification Using Quadric Error Bounds”,
ACM Proc. Siggraph '
97, August 1997, pp. 209-216, and ordering based on visibility, as reported by D. Aliaga in “Visualization of Complex Models Using Dynamic Texture-Based Simplification”,
Proc. IEEE Visualization '
96, October 1996, pp. 101-106, all of which are targeted towards minimizing the delay before the client is able to generate an initial display. However, producing such representations may involve significant server computing and storage resources, the downloading time remains large for complex models, and additional time may be necessary on the client to process the data received (e.g., decompression). For example, Adaptive Media's Envision 3D (see www.envision.com) combines computer graphics visibility techniques (e.g., occlusion culling as described by H. Zang et al., “Visibility Culling Using Hierarchical Occlusion Maps”,
ACM Proc. Siggraph '
97, August 1997, pp. 77-88) with streaming to guide the downloading process by sending to the clients the visible geometry first and displaying it as it is received, rather than waiting for the entire model to be sent. Nonetheless, determining which geometry is visible from a given viewpoint is not a trivial computation and maintaining acceptable performance remains a challenging proposition even when only visible geometry is transmitted.
At the opposite end of the spectrum are server-side rendering methods, as generally shown in
FIG. 3
, which place the burden of rendering a model entirely on the server and the images generated are subsequently transmitted to clients. As in the case of client-side methods, the client
301
usually initiates a request for a model. However, instead of downloading the three-dimensional model to the client
301
, the model and scene description
302
stored in storage
303
is rendered on the server
304
in rendering engine
305
to produce two-dimensional static images
306
, and one or more two-dimensional images
307
resulting from this rendering are transmitted over the network
308
to the client
301
. Subsequently, the images
307
are displayed on display
309
of the client
301
. The cycle is then repeated based on user feedback
310
.
Such techniques have the advantages that they do not require any three-dimensional graphics capabilities on the part of the clients and the bandwidth requirements are significantly reduced. The tradeoffs in this case are the loss of real-time interaction with the model (i.e., images cannot be delivered to clients at interactive frame rates) and the increase in server load and hence, server response times, as the number of clients concurrently accessing the server increases. An example of a server-side-based rendering system is CATWeb (www.catia.ibm.com) which is a web browser-based application designed to provide dynamic CAD data access to users with intranet connections and graphics capabilities. Another example in this category is panoramic rendering described by W. Luken et al. in “PanoramIX: Photorealistic Multimedia 3D Scenery”,
IBM Research Report
#RC21145, IBM T. J. Watson Research Center, 1998. A panorama is a 360 degree image of a scene around a particular viewpoint. Several panoramas can be created for different viewpoints in the scene and connected to support limited viewpoint selection.
Hybrid rendering methods described by D. Aliaga and A. Lastra in “Architectural Walkthroughs Using Portal Textures”,
Proc. IEEE Visualization '
97, October 1997, pp. 355-362, M. Levoy in “Polygon-Assisted JPEG and MPEG Compression of Synthetic Images”,
ACM Proc. Siggraph &ap
Borrel Paul
Hall Shawn
Horn William P.
Klosowski James T.
Luken William L.
Percello Louis J.
Vo Cliff N.
Whitham Curtis & Christofferson, P.C.
LandOfFree
Method and apparatus for delivering 3D graphics in a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for delivering 3D graphics in a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for delivering 3D graphics in a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2885784