Workshop "Ubiquitous Tracking", Salzburg 06-08 Feb 2004
protocol 08 Feb 2004: internal discussion about diploma theses
Attendance
Dagmar Beyer, Daniel Pustka, Franz Strasser
thematical separation of the diploma theses
- Daniel: application for calibration, fusion components
- Franz: connecting subgraphs, independent from calibration and data flow graphs
- Dagmar: path search on spatial realtionship graphs
constistent IDs for objects, no consideration of anonymous objects
dependencies
- Franz - Dagmar: both must work on the spatial relationship graphs.
find a suitable representation in DWARF
- Daniel - Franz/Dagmar: calibration needs at least two different paths between two objects.
could be selected manually for testing
further considerations
- Franz:
- Mapping the topology of the Spatial Relationship Graph and the logical topology of the underlying network
- Connecting different logical network subsystems
- Connecting two differnt SR-Graphs (SR1, SR2)
observe spatial relation between nodes in SR1 and SR2 in order to find connecting edges between SR1 and SR2
- classification of new objects or anonymous measurements:
- anonymous object in SR1: new object or anonymous object (measurement of a motion detector for example)
- already identified object in SR2
- same motion modell, same position: insert an ID-edge between them
- Dagmar:
- evaluation of the error function
- inferred attributes for Data Flow Components such as Interpolation Component, etc
- implementation and anlaysis of different path-search algorithms for different classes of error functions
- complexity analysis
- experimental results in simulation environment
Agenda
- list requirements for the diploma theses
- set focus on main problems
- optimisation only in those parts
- priorisation of tasks/implementations
- find representation of the spatial relationship graph in DWARF
- definition of error function and attribute mapping
- how far is fusion (loosely coupled) part of the Ubitrack system?
Agenda
--
DagmarBeyer? - 17 Feb 2004
