No abstract

With tracking setups becoming increasingly complex, it gets more difficult to find suitable algorithms for tracking, calibration and sensor fusion. A large number of solutions exists in the literature for various combinations of sensors, however, no development methodology is available for systematic analysis of tracking setups.When modeling a system as a spatial relationship graph (SRG), which describes coordinate systems and known transformations, all algorithms used for tracking and calibration correspond to certain patterns in the graph. This paper introduces a formal model for representing such spatial relationship patterns and presents a small catalog of patterns frequently used in augmented reality systems. We also describe an algorithm to identify patterns in SRGs at runtime for automatic construction of data flows networks for tracking and calibration.

No abstract

Ubiquitous tracking setups, covering large tracking areas with many heterogeneous sensors of varying accuracy, require dedicated middleware to facilitate development of stationary and mobile applications by providing a simple interface and encapsulating the details of sensing, calibration and sensor fusion.In this paper we present a centrally coordinated peer-to-peer architecture for ubiquitous tracking, where a server computes optimal data flow configurations for sensor and application clients, which are directly exchanging tracking data with low latency using a lightweight data flow framework. The server's decisions are inferred from an actively maintained central spatial relationship graph using spatial relationship patterns.The system is compared to a previous Ubitrack implementation using the highly distributed DWARF middleware. It exhibits significantly better performance in a reference scenario. MOTIVATIONIn industrial augmented reality scenarios, there is a growing demand for integrated working environments which span large factory buildings. In such an environment, many different mobile and stationary AR-supported applications, such as logistics, production, maintenance or factory planning may coexist and require shared access to permanent tracking with varying accuracy requirements. Today, no single technology exists that satisfies the tracking requirements of all these applications and can -at least for a reasonable price -be deployed throughout such an environment. For this reason, in a realistic setup, many different tracking systems would be installed ranging from low-precision wide-area WLAN tracking to infrared-optical systems covering only small areas with high accuracy. The installation, maintenance and expansion of such a largescale heterogeneous tracking environment poses new challenges to the underlying middleware concepts.Heterogeneous wide-area tracking environments Emerging tracking methods based on technologies like WLAN or RFID provide the possibility to deploy tracking to ever-enlarging indoor areas. With increasing tracker coverage, a larger diversity of AR applications will need to share this tracking infrastructure. Stationary applications that are already in use will more and more be complemented by mobile applications that would have been completely impossible without wide-area tracking. Also, applications that are stationary today, might benefit from enlarging tracking areas and * e-mail: { huberma, pustka, keitler, echtler, klinker }@in.tum.de become more adaptive and better integrated in the productive environment. Many of these wide-area tracking systems have the drawback of being rather imprecise. Nevertheless, they serve quite well for navigation problems and can thereby bridge the gap between islands of higher tracking accuracy. Furthermore, they can provide useful initial positions to other sensors, such as markerless optical trackers [7]. There are also many examples where a fusion of measurements from different mobile and stationary sensors improves overall trac...

Marker-based optical tracking systems are widely used in augmented reality, medical navigation and industrial applications. We propose a model for the prediction of the target registration error (TRE) in these kinds of tracking systems by estimating the fiducial location error (FLE) from two-dimensional errors on the image plane and propagating that error to a given point of interest. We have designed a set of experiments in order to estimate the actual parameters of the model for any given tracking system. We present the results of a study which we used to demonstrate the effect of different sources of error. The method is applied to real applications to show the usefulness for any kind of augmented reality system. We also present a set of tools that can be used to visualize the accuracy at design time.

No abstract

No abstract

In this paper, we present a mobile augmented reality application that is based on the acquisition of user-generated content obtained by 3D snapshotting. To take a 3D snapshot of an arbitrary object, a point cloud is reconstructed from multiple photographs taken by a mobile phone. From this, a textured polygon model is computed automatically. Other users can view the 3D object in the environment of their choosing by superimposing it on the live video taken by the cell phone camera. Optical square markers provide the anchor for virtual objects in the scene. To extend the viewable range and to improve overall tracking performance, a novel approach based on pixel flow is used to recover the orientation of the phone. This dual tracking approach also allows for a new single-button user interface metaphor for moving virtual objects in the scene. The Development of the AR viewer was accompanied by user studies and a further summative study evaluates the result, confirming our chosen approach.Keywords: Mobile augmented reality, 3D snapshotting, tracking, user interface MotivationWe present a mobile augmented reality (AR) platform based on user-generated content. The core idea is to enable a user of our system to generate a 3D model of arbitrary small or mid-sized objects, based on photographs taken with their mobile phone camera. Thereupon, another user can inspect the object integrated in their natural environment, e.g. their home, using our mobile AR viewer application. We refer to this capture and viewing process as "3D Snapshotting" 3D models can be generated on-the-fly, using a pair of photographs of an object from different perspectives and reconstruct its 3D structure from them. This results in a dense 3D point cloud.