Dipl-Wirtsch.-Ing. Marco Moebus
Member of the Signal Processing Group at the Institute of Telecommunications, TU Darmstadt.
Office: S3 / 06 252
work +49 6151 16-2306
fax +49 6151 16-3778
Marco received his diploma in electrical engineering and management (Dipl.-Wirtsch.-Ing.) from Technische Universität Darmstadt in 2005. His diploma thesis in the area of speech recognition was conducted within the ASIMO project at the Honda Research Institute Europe. In April 2006, Marco started working on his PhD in the Signal Processing Group.
2D Sparse Array Design
Research is conducted in the field of array signal processing, with an emphasis on the design of 2D/3D sparse arrays. These nonuniform arrays allow equal or sometimes better performance than conventional arrays with considerably less hardware/computational effort. Design principles can be motivated using nonuniform sampling theory.
While applications arise from sonar, special cases of mobile communications and imaging problems, we focus mainly on automotive safety and comfort applications, robot navigation and other scene reconstruction scenarios. Here, the focus is mainly in the development of new sensor systems based on ultrasound sensor arrays.
Object Detection and Classification Using Acoustic Cameras
Using 2D ultrasound arrays, it is possible to visualize and locate objects three-dimensionally in the surroundings of a car, robot or any other mobile platform. Furthermore, we work on the problem of object classification based on the obtained acoustic images, e.g. to detect the presence of humans for attention control and subsequent actions. An acoustic camera allows a very reliable representation of important features of the surrounding, independent of lighting conditions and largely indifferent to properties of the object's material. This makes those cameras a cheap and reliable sensor alternative or augmentation to light- or radar based short-range systems.
- Digital Signal Processing (2006-2009)
Local Co-Chair of “The Fifth IEEE Sensor Array and Multichannel Signal Processing Workshop” (SAM), Darmstadt, July 2008. Website
Student Projects (SA/DA/BT/MT)
In the context of above mentioned research work, I do supervise student projects in the area of acoustic imaging and modelling as well as array signal processing for coherent sources. The following list represents a selection of topics and can be adapted to the profile of the student. It is also sometimes possible to work an a specific topic as a team of students.
- based on Minimum-Redundancy-Array-Theory
- with respect to US sensor characteristics
- in the context of scanning in imaging systems
- 3D-Obstacle-Detection with Ultrasound array
- Object classification by Modelling physical Parameters
- Pedestrian Detection by Feature Analysis of Echo signals
- Imaging based on Subarray Techniques
Prerequisites are motivation for the topic, a general interest in methods and concepts of array signal processing and the willingness to work on a topic under own responsibility. If you have any questions, just come to my office.
Completed student thesis projects
Gerhard, Bernd Feature Analysis and Classiﬁcation of Walking and Standing Humans Using Ultrasound Arrays
Fang, Xiyue Redundancy Analysis of Nonuniform 2D Sensor Arrays
Degenhardt, Holger Calibration Methods for 2D Arrays
Fort, Carles Weighting Optimization for 2D Sparse Array Beamforming
Jennewein, Dirk Array Design Optimization Using Evolutionary Algorithms
Ngouoko, Samuel Design Optimization of 2D Sparse Arrays
Pötzl, Katja Pedestrian Detection for Ultrasound Images
Qin, Lu Image Fusion Methods for Ultraound Imaging
Withopf, Jochen Detection of Acoustical Footstep Signatures