Research

My research is mainly focused on vision systems for mobile platforms.

I am currently working on image stabilisation of virtual cameras extracted from omnidirectional video using a probablisic approach and sensor integration with an inertial measurement unit. This approach has been applied to multi view tracking in maritime environments:

• Image Stabilisation
• Multi View Tracking
• Tracking/Clustering on the Sphere

Furthermore, I am working on saliency inspired visual attention. I focus on the use of multiple low-level features to estimate areas of visual attention in images and omnidirectional videos captured by maritime platforms. This work has been further extended by using different metrics for feature evaluation as well as a domain specific background segmentation technique:

• Maritime Saliency

I am involved in a research project for designing domain specific languages for client-server applications. See this video about an example application that uses device independent state charts for code generation of user interfaces for iOS and Android devices, as well as automatic protocol generation for transmission of motion sensors and video data:

• Multi-Client Remote Viewing of Omnidirectional Video Using Virtual Cameras

My previous research about Indoor Localisation and Sensor Fusion with RFID and Computer Vision has been carried out at Dortmund University of Applied Sciences. Please visit the pages of the Institute of Process Improvement & CAQ or contact Burkhard Igel and Ulf Niemeyer if you are interested in their current projects.