The group named Tracker-algorithms and Tracker-assessment addresses the development, implementation, and application as well as assessment of algorithms for object tracking based on image sequences recorded by passive imaging sensors. A large variety of objects, platforms, sensors, object ranges, and backgrounds are considered. This includes flying objects, surface vehicles and ships as targets as well as platforms, sensors for the visual and infrared wavelengths, object ranges from meters up to several kilometers and unstructured backgrounds as clear sky and open sea as well as structured backgrounds like in urban environments.
A growing need to assess systems with included tracker-algorithms is observed. This assessment can be performed by an absolute comparison of the system output with ground-truth-data as well as by comparison of different systems. Different assessment schemes and different assessment criteria have to be applied dependent on the usage and the context of the systems. In our group, we operate different systems for ground-truth-data acquisition and processing as well as programs for the automatic comparison and assessment of system output data with ground-truth-data.
Dr. Leo Doktorski
Dr. Eckart Michaelsen
Dr. Thomas Perschke
Dr. Norbert Scherer-Negenborn
- Distributed and adaptive multisensor fusion engine (DAFNE)
Project DAFNE created software for sensor data fusion improving situations awareness. It was a multinational EDA project with contributions from Italy, Netherlands, Poland, Czech Republic, and Germany running from 2009 to 2011. Particular focus was the use of context information and assessment of earlier data into the weighted fusion. IOSB has provided comparison between results and ground-truth and assessment of the gain.
- Naval target recognition and protection
This project aims at evaluating possible countermeasures of a ship against thermal seeker guided weapons. Generic variants of trackers and their performance as function of environment and countermeasures are investigated and assessed, resulting in threat ranges and hit-probabilities. In the end this will help the captain with decisions concerning countermeasures and keeping safe distances.
- Experimental setup for object recognition and tracking
For the evaluation of novel imaging sensors, hardware architectures and image processing algorithms we are providing a real-time object recognition and tracking system. The modular design of the systems allows for an easy exchange and upgrade of the components. The system is also used for the development of new implementation concepts. One example is an multi-object tracker utilizing the dynamic, partial reconfiguration capability of several FPGAs. We provide system adaptions to use it in outdoor/maritime environments.
- New concepts for the programming of heterogeneous hardware
Many systems ranging from desktop computers to smartphones use different hardware architectures like processor cores and GPUs, forming a heterogeneous hardware system. The efficient utilization of these different available components is a new challenge for software developers. We investigate new concepts for an easy and efficient programming of such heterogeneous systems.
- Measurement of vibrations on rotating wind energy plants
Rising numbers of wind energy parks give new challenges concerning maintenance and damage diagnosis. Measuring the wings’ vibrations in operating mode can be a good option. IOSB is constructing a demonstrator capable of doing so. From a distance of about 300m a laser-spot is controlled to cm accuracy on a position of choice along the wing.
Inspection and documentation of underwater infra-structure (e.g. harbor moles, bases of bridges, or offshore wind energy plants) require the effort of professional diving crews, which is expensive and sometimes dangerous. An automation of image acquisition of such buildings increases the operation time and the repeatability even under worse weather conditions. For off-line analysis the recorded images are stitched to large panoramic underwater views. The needed algorithms for creation, visualization and comparison of such panoramic views are developed, implemented and tested.