DETAILS


COLUMNS


CONTRIBUTIONS


ARCHIVE



FROM THE EDITOR


Vol.33 No.4 Nov. 1999
ACM SIGGRAPH


Explore the Fascinating Fusion of Computer Vision and Computer Graphics




Gordon Cameron
SOFTIMAGE, Inc.

November Columns
About the Cover

From the Editor
Previous From the Editor Previous From the Editor

The copyright of articles and images printed remains with the author unless otherwise indicated.

The field of computer graphics continues to innovate and advance at an astonishing rate, and I’m delighted that we’re able to focus the spotlight this issue on one particularly active area.

Eight or so years ago, I worked briefly in a university lab as a research assistant in computer vision and robotics. At that time there seemed a clear dividing line between those two disciplines and computer graphics. Sure - one of my co-workers had developed a modeler and visualizer for showing shaded models that we were using for matches against laser scanned objects, and I’d worked on some simple graphic displays for visualizing matches and decisions trees, but that was about the extent of the overlap at our own facility. Computer graphics, for me, meant scientific visualization, or The Abyss, Terminator 2 or Luxo Jr. We were not doing anything even vaguely related to the movies - or so I thought at the time.

After leaving that position and moving to work directly with computer graphics, however, I started to notice an astonishing crossover of research (and researchers!) - the words ‘computer vision’ were cropping up increasingly in the literature and SIGGRAPH papers, panels and courses. The last year or so has seen an explosion of interest in the area (and perhaps even a general public awareness with the release of computer graphics/vision effects-heavy motion picture The Matrix), and a focus in Computer Graphics seemed timely.

IBR, vision for interaction, motion editing and capture, camera tracking and stabilization are all examples of this exciting fusion. For this issue’s theme exploring these and other examples, I was lucky enough to have Rick Szeliski and Steve Seitz (well respected researchers spanning disciplines) agree to guest edit and gather a fantastic selection of authors and articles. Together, the papers present a fascinating picture of the increasing utilization of computer vision techniques in computer graphics.

After reading the issue, I have no doubt that the next decade will yield remarkable advances in visual realism and imagery in general, thanks in great part to this relatively new and exciting coupling of disciplines. That expensive hand built laser scanner we used in the vision lab may soon seem rather quaint, what with the recent range movie camera by 3DV Systems, the ZCAM [1] (allowing more, one hopes, than just the real-time compositing of frustrating computer graphics animated characters into frustratingly fast-paced music shows with frustrated and frustrating presenters!), and with the availability of new off-the-shelf consumer handheld scanners or cameras claiming to help build 3D models (such as Minolta’s 3D 1500 [2]), coupled with software like Metacreations Canoma [3], which helps build such models from standard photographs (models that perhaps I can send to my Mum’s personal 3D fax machine any year now!).

We’re starting to come full circle as this year at SIGGRAPH I managed to meet up with that co-worker from my vision days who is now an active researcher in the computer graphics field and contributing directly to an area that will doubtless have a direct impact on the visuals in the next The Matrix. Who would have thought?!

As ever, my thanks go out to all those involved in helping with the issue. I welcome your comments and feedback, and will see you again in February when we turn our attention to scientific visualization.

References

  1.  See http://www.3dvsystems.com/.
  2.  See http://www.minolta.com/dp/3d1500/index.html.
  3.  See http://www.metacreations.com/products/canoma/ and http://www.canoma.com/.


Computer Graphics Editor Gordon Cameron spent his childhood skimming stones at the beach in Banff, Scotland and did his first programming at age 13 on a kindly neighbour’s Apple II Europlus. In 1995, he started working for SOFTIMAGE, Inc. in Montreal where he currently acts as Project Leader for Animation in the 3D team.

Gordon Cameron
Software Development
SOFTIMAGE, Inc.
3510 boul. St-Laurent
Suite 400
Montreal, Quebec
H2X 2V2
Canada

Tel: +1-514-845-1636 ext.3445
Fax: +1-514-845-5676