TeachingSs06HSIndustrialAR

Chair for Computer Aided Medical Procedures & Augmented Reality
Lehrstuhl für Informatikanwendungen in der Medizin & Augmented Reality

Applications of Computer Vision for Industrial Augmented Reality

2 SWS
Wahlfach
ECTS: 4 Credits

The course will be conducted in English.

ALERT! The Haupseminar is moved to Monday 24/07/2006. It will be in the room 03.13.010.

Begin: April 2006, 2 hours per week.

Registration

Please send an email to
Please include last and first name, matriculation number, current semester, discipline (diploma, master, bachelor) including major, and a desired topic.

Introduction and Abstract

First applications of industrial Augmented Reality (AR) were proposed, designed and developed in the early nineties. As the high expectations for rapid progress and easy adoption of virtual reality (VR) solutions within industrial environment began to fade, AR could be seen as a possible remedy. Advances in wireless communication, global positioning, ubiquitous and wearable computing, as well as optical and video see-through display technologies, supported the idea of augmenting the industrial users rather than forcing them to immerse into an unfamiliar virtual environment. The main hope was to reduce the costs for (a) building comprehensive virtual models of the world, (b) developing new user interaction paradigms, and (c) training the industrial users to work in a totally new environment.

mini-ar-sm.jpg which_is_real-sm_small.jpg


Topics

Topic Detailed Description Literature Student Tutor Handouts/Presentations
Template-based tracking 1 Since the Lucas-Kanade algorithm was proposed in 1981 image alignment has become one of the most widely used techniques in computer vision.
Applications range from optical flow and tracking to layered motion, mosaicing, and face coding.
* Lucas, B. and Kanade, T. (1981). «An iterative image registration technique with application to stereo vision». In Int. Joint Conf. on Artificial Intelligence, pages 674–679.
* Shi, J. and Tomasi, C. (1994). «Good Features to Track». In IEEE Int. Conf. on Computer Vision and Pattern Recognition, pages 593–600.
* Hager, G. and Belhumeur, P. (1998). «Efficient region tracking with parametric models of geometry and illumination». IEEE Trans. on Pattern Analysis and Machine Intelligence, 20(10):1025–1039.
  Selim BenHimane  
Template-based tracking 2 Numerous algorithms have been proposed and a wide variety of extensions have been made to the original formulation of the Lucas-Kanade algorithm. * Baker, S. and Matthews, I. (2004). «Lucas-Kanade 20 years on : a unifying framework». International Journal of Computer Vision, 56(3):221–255. Kenan Bektas Selim BenHimane 24.7. 13:00 in 03.13.010
Feature-based matching For augumented reality applications, one has to find the relative position between the camera and the scene. One common approach for doing this is by combining feature point descriptors with a fast classifier to obtain real-time detection. This detection makes it possible to augument even deformable objects in real-time. * V. Lepetit, P. Lagger and P. Fua, Randomized Trees for Real-Time Keypoint Recognition, Conference on Computer Vision and Pattern Recognition, San Diego, CA, June 2005
* J. Pilet, V. Lepetit and P. Fua, Real-Time Non-Rigid Surface Detection, Conference on Computer Vision and Pattern Recognition, San Diego, CA, June 2005,
* J. Pilet, V. Lepetit and P. Fua, Augmenting Deformable Objects in Real-Time, International Symposium on Mixed and Augmented Reality, Vienna, Austria, October 2005
  Andreas Hofhauser  
Feature-based tracking and sensor fusion Rapid camera rotations (e.g. camera shake) are a significant problem when real-time computer vision algorithms are applied to video from a handheld or head-mounted camera. Such camera motions cause image features to move large distances in the image and cause significant motion blur.
A fast method of estimating the camera rotation from a single frame which does not require any detection, matching or extraction of feature points is needed and can be used as a motion estimator to reduce the search range for feature matching algorithms that may be subsequently applied to the image.
* Drummond, T. and Cipolla, R. (2002). «Real-time visual tracking of complex structures». IEEE Trans. on Pattern Analysis and Machine Intelligence, 24(7) :932–946.
* Klein, G and Drummond, T. (2002), «Tightly Integrated Sensor Fusion for Robust Visual Tracking.» BMVC.
Syed Omar Farooq Pierre Georgel  
Hybrid tracking Feature-based tracking and template based-tracking can be mixed in order to make a more robust tracking system. It can be done sequentially (Bascle et al., 1994; Chiba and Kanade, 1998; Marchand et al., 2001b) or simultanously (Masson et al., 2004a; Vacchetti et al., 2004; Pressigout and Marchand, 2005). * Bascle, B., Bouthemy, P., Deriche, R., and Meyer, F. (1994). «Tracking complex
primitives in an image sequence». In Int. Conf. on Pattern Recognition.
* Chiba, N. and Kanade, T. (1998). «A Tracker for Broken and Closely Spaced Lines». In Proceedings of the 1996 International Society for Photogrammetry and Remote Sensing Conference, pages 676 – 683.
* Marchand, E., Bouthemy, P., and Chaumette, F. (2001b). «A 2D-3D model-based approach to real-time visual tracking». Image and Vision Computing, 19(13) :941–955.
* Masson, L., Dhome, M., and Jurie, F. (2004a). « Robust Real Time Tracking of 3D Objects ». In Int. Conf. on Pattern Recognition, pages 252–255.
* Vacchetti, L., Lepetit, V., and Fua, P. (2004). «Combining Edge and Texture Information for Real-Time Accurate 3D Camera Tracking». In 3rd IEEE and ACM International Symposium on Mixed and Augmented Reality, pages 48–57.
* Pressigout, M. and Marchand, E. (2005). « Real-time planar structure tracking for visual servoing : a contour and texture approach ». In IEEE/RSJ Int. Conf. on Intelligent Robots and Systems.
Jiran Wang Selim BenHimane 24.7. 14:00 in 03.13.010

Schein requirements

As usual:
  • attendance of the weekly meeting (one time absence allowed, otherwise we'll request a medical certificate or some other very good reason)
  • active class participation (discussions)
  • presentation in the weekly meeting (approximately 60 minutes + discussion)

Timeline

You are required to meet the following deadlines for discussing your work with your supervisor:
  • 4-6 weeks before presentation: Outline
  • 2 weeks before presentation: Handout
  • 1 week before presentation: Presentation Slides

Templates


TeachingForm
Title: Applications of Computer Vision for Industrial Augmented Reality
Professor: Nassir Navab
Tutors: Selim Benhimane and Pierre Georgel and Andreas Hofhauser
Type: Hauptseminar
Information:  
Term: 2006SoSe
Abstract: First applications of industrial Augmented Reality (AR) were proposed, designed and developed in the early nineties. As the high expectations for rapid progress and easy adoption of virtual reality (VR) solutions within industrial environment began to fade, AR could be seen as a possible remedy. Advances in wireless communication, global positioning, ubiquitous and wearable computing, as well as optical and video see-through display technologies, supported the idea of augmenting the industrial users rather than forcing them to immerse into an unfamiliar virtual environment. The main hope was to reduce the costs for (a) building comprehensive virtual models of the world, (b) developing new user interaction paradigms, and (c) training the industrial users to work in a totally new environment.


Edit | Attach | Refresh | Diffs | More | Revision r1.34 - 26 Jan 2007 - 17:17 - MartinHorn

Lehrstuhl für Computer Aided Medical Procedures & Augmented Reality    rss.gif