Recheneffiziente Algorithmen zur Objektidentifikation für materialdurchdringendes UWB-Radar

Sebastian Hantscher

Research output: ThesisDoctoral thesis

Abstract

The goal of this research project is the development of concepts for the identification of targets in or behind materials, in particular the detection of water pipes in walls. Initially, a method for the extraction of characteristic feature of a target was proposed. For this purpose a Wiener filter based radar calibration procedure was implemented enabling the estimation of the frequency-dependent radar cross section only using A-scans. Another focus lied on the application of the concepts of the synthetic aperture for pulse based radar in order to reconstruct the surface of the objects. On the other hand, the calculation time was a main point considered during the development of appropriate algorithms. It led to signal processing which is outclass to conventional methods. They were verified by real radar data recorded by the UWB hardware developed at the institute. A key feature of imaging radar systems is the achievable resolution, describing the ability to distinguish between two targets being close to each other. Because it is defined by the used bandwidth as well as the antenna und therefore system inherent, the goal was to enhance the resolution by appropriate signal processing methods. This led to pulse separation algorithms. Two types of separations methods were developed: correlation based methods and a method based on genetic optimisation. The functionality of the algorithms were demonstrated on a measurement of the water pipes in a gas concrete wall. The pipes were detected and imaged although their echoes were not distinguishable by the raw data.
Original languageGerman (Austria)
Publication statusPublished - Oct 2008

Fields of science

  • 202019 High frequency engineering
  • 202028 Microelectronics
  • 202039 Theoretical electrical engineering
  • 202037 Signal processing
  • 202027 Mechatronics
  • 202036 Sensor systems

JKU Focus areas

  • Mechatronics and Information Processing

Cite this