Winter Augmented Reality Meeting 2005
19th December 2005, Planner Alm, Austria
The fields of Computer Graphics, Augmented Reality, Computer Vision and Ubiquitous Computing have the potential to be synergistic. However, the overlap and mutual contributions of each area has yet to be expressed and understood. The specialist, immersed in his or her own discipline, struggles to see the "big picture". The core of the workshop is to be Augmented Reality and we are soliciting presentations from members of each community that illuminate the interplay between the different disciplines and AR.
The original Call for Participation
9:00 - 10:15
- Gerhard Schall - Experiences in Building 3D Indoor Models for Augmented Reality
In general there are two techniques a spatial 3D model can be created. Manual surveying techniques involve the use of Total stations to survey the positions of fiducial markers. This manual approach is time-consuming, and presents a serious barrier to the introduction of AR to new environments. Therefore automatic methods are necessary to speed up the process. We use an autonomously navigating mobile robot equipped with a laser range finder and a camera to detect and localise the fiducial markers and build a 3D model. Nevertheless, by performing a high quality survey of fiducial markers and their environs using the most sophisticated geodetic equipment, we can establish a ground truth against which vision techniques can be evaluated. The manual surveying process workflow will be discussed with its pros and cons as well as experiments with the autonomous navigating robot. Additionally a technique that combines manual and automated approaches to creating and maintaining an environment of AR and Ubicomp applications is shown.
- Alexander Bornik - A Hybrid User Interface for Liver Surgery Planning
A novel system for interactive visualization and manipulation of medical datasets in the context of surgery planning based on a hybrid VR / Tablet PC user interface will be presented. The goal of the system is to facilitate efficient visual inspection and correction of surface models generated by automated segmentation algorithms based on x-ray computed tomography (CT) scans, needed for planning surgical resections of liver tumors. Factors like the quality of the visualization, nature of the dataset and interaction efficiency strongly influence system design decisions, in particular the design of the user interface, input devices and interaction techniques, leading to a hybrid setup.
- Daniel Wagner - Handheld AR
Current developments - Challenges - Future plans
10:45 - 12:00
- Denis Kalkofen - OpenVideo - A Video Abstraction Layer
OpenVideo is a general data integration and processing software with special support for video data. It implements a hardware abstraction layer for video devices by interfacing several different device drivers either directly or through the functionality of third party video libraries. OpenVideo is designed to be as extensible and easily configurable as possible. This talk will give an overview of OpenVideo's design concepts and its data structure. The current set of OpenVideo's functionality and its concept of extending this set will be demonstrated and finally an overview of the implemented concepts of integrating OpenVideo's data into other programs will be shown.
- Margrit Gelautz - Video Object Segmentation and Compositing
In this presentation, we focus on techniques for video object segmentation and the subsequent combination of the extracted object(s) with another video scene ('video compositing') from a computer vision point of view. We present results obtained by state-of-the art stereo and motion segmentation algorithms and discuss the special requirements of video compositing tasks on the segmentation algorithm. In this context, we also address special problems arising along object boundaries and related occlusions and demonstrate the need for boundary matting techniques.
- Gerhard Reitmayr - Augmented Reality without models
14:00 - 15:15
- Tom Drummond - Are user interfaces in the world AR?
- Joe Newman - AR out of the box - Five years on
15:45 - 17:00
Brainstorming session: Briding the Visual Gap
Papers at ISMAR typically fall into two categories: Presenting AR results that do not use Computer Vision or presenting Computer Vision results that use AR as motivation. Computer Vision continues to promise improvements to AR user interfaces on all levels. Therefore, we should think about a transition of recent results into (existing) AR systems. Such integration will foster new research directions in both fields. In this session we would like to think about possible projects that we could undertake by combining results from participants of the WARM'05 workshop.