Christian Weiss

Sensor array based on split ring resonators for analysis of organic tissues

A sensor array concept has been developed using microstrip-line-excited split ring resonators (SRR). With the proposed structure it is possible to spatially resolve the dielectric properties of a Material Under Test (MUT). The split rings are designed to have different resonant frequencies and are decoupled from each other to allow a spatial distribution where a frequency shift of one individual resonant peak will indicate the dielectric properties of the MUT and its location within the array. Several prototype sensors have been realized and tested with different MUT such as dielectric bricks and pig lung tissue to prove the concept.

DOA estimation in the presence of array imperfections: A sparse regularization parameter selection problem

A robust sparse regularization technique for source localization that accounts for the joint effects of sensor position errors and noise is presented. Finding a good choice of the regularization parameter is a key component in sparse optimization problems and its automated determination is typically a non-trivial task. Our approach attempts to statistically determine an upper bound of the mean-squared error resulting from noise and from uncertainty about the exact sensor positions. Hereby, we aim at finding a direct relation between the physical parameters of the array, i.e. the sensor position errors, and the hyperparameter in the constrained formulation of the optimization problem. We will show that the proposed method provides proper sparse regularization even in low SNR regimes and in the presence of severe array imperfections.

Distributed target detection in Through-the-Wall Radar Imaging using the bootstrap

The problem of distributed detection and decision fusion in Through-the-Wall Radar Imaging (TWRI) is considered. We deal with the multi-viewing case in which images corresponding to different radar locations can be collected. We present a method to adapt conventional distributed detection schemes to the scenario when no a priori knowledge about image statistics from any view is available. Further, a new scheme for estimating quality information of local detectors in a distributed detection scenario is proposed. We apply bootstrap techniques to draw inference from the radar measurements of the behind the wall scene. Simulation results as well as experimental data are used to demonstrate the performance of the proposed approach.

Compressed sampling and dictionary learning framework for wavelength-division-multiplexing-based distributed fiber sensing

We propose a compressed sampling and dictionary learning framework for fiber-optic sensing using wavelength-tunable lasers. A redundant dictionary is generated from a model for the reflected sensor signal. Imperfect prior knowledge is considered in terms of uncertain local and global parameters. To estimate a sparse representation and the dictionary parameters, we present an alternating minimization algorithm that is equipped with a preprocessing routine to handle dictionary coherence. The support of the obtained sparse signal indicates the reflection delays, which can be used to measure impairments along the sensing fiber. The performance is evaluated by simulations and experimental data for a fiber sensor system with common core architecture.

A sparse regularization technique for source localization with non-uniform sensor gain

A robust sparse regularization technique with applications to source localization in the presence of non-uniform gain distribution is presented. As a key component in sparse optimization, a proper choice of the regularization parameter is crucial. It renders high impact on the localization performance and has to account for any kind of model perturbation. Our proposed technique utilizes a statistical framework to provide a direct relation between the physical system parameters and the regularization parameter of the resulting optimization problem. It addresses the joint effects of sensor gain variations and noise. As a figure of merit, we consider the mean-squared error (MSE) between the perturbed measurements and the assumed underlying model. An upper bound of the MSE is attained in order to estimate the regularization parameter. The presented method shows good performance for moderate gain variances even in low SNR regimes.

How good is your Super-Resolution Image? Quality assurance in image reconstruction using the bootstrap

Super-Resolution Image Reconstruction is known to be sensitive to errors in assumptions such as accurate sub-pixel motion estimation. Even small errors can yield a significant degradation of image quality that complicates any follow-on task such as object detection or classification. We focus on the problem of automatic quality assessment of Super-Resolution image reconstruction. We propose a bootstrap-based method that provides an objective metric quantifying reconstruction quality and thus allowing to readjust the reconstruction.