Optical smoke detectors are designed to detect small concentrations of smoke to ensure a fast and reliable detection of arising fires. Unfortunately the complex task of avoiding false alarms is not completely addressed. In contrast to the well standardized methods for the evaluation of the detection capability of a smoke detector, there is a lack of reproducible and representative test methods concerning the false alarm susceptibility with regard to nuisance aerosols. A recent study says that about 10 % of false alarms are caused by dust. For this reason this paper presents a new approach for the test of optical smoke detectors regarding their susceptibility to false alarms due to the nuisance aerosol dust. The presented test apparatus is a very helpful and important tool for developers as well as for system designers having a quantitative decision criterion to find the optimal detector for a specific scenario.

Terahertz time-domain systems are known as a precise imaging tool. These systems make use of parabolic mirrors or lenses to illuminate a small spot of a sample under test. By moving the sample or the sensor head, an image can be recorded. This imaging technique guarantees a high signal-to-noise ratio and large bandwidth. However, using this method a priori knowledge of the sample shape is needed for the correct focusing of the system. This limits the performance and robustness of the imaging system as only specular reflections are considered for the image. Here, we propose a fast reflection-based broadband terahertz time-domain imaging method that overcomes these hurdles by making use of both the specular and diffuse reflections of a divergent terahertz beam. The proposed method employs no optical lenses or mirrors but uses signal processing and classical radar migration techniques. High-resolution imaging is achieved by focusing the divergent terahertz beam via post-processing. To compensate the inherent poor signal-to-noise ratio of the unfocused terahertz beam, calibration and post-processing methods are used. For the evaluation of the imaging method, geometrically complex samples are scanned by a fast terahertz time-domain spectroscopy system based on electronically controlled optical sampling. The bandwidth achieved with the divergent beam is 2.5 THz with a signal-to-noise ratio of around 30 dB. We demonstrate that this method is capable to generate high-resolution 2D terahertz images of objects with arbitrary size, shape, orientation and relative position to the emitter and detector antennas. Objects with sub-mm dimension can be clearly reconstructed for arbitrary positions and orientation achieving resolution in the µm region. Furthermore, the presented method can be applied for any reflection-based scenarios and antenna configuration.

For the purpose of a high-precision object recognition (OR) system at THz frequencies, a 140 GHz super-resolution imaging system is presented in this paper. Conventional THz camera imaging systems are limited by their pixel numbers resulting in distorted and noisy low-resolution (LR) images hindering the possibility of correct object recognition. In this paper methods for improving the resolution of a 140 GHz camera are proposed based on super-resolution (SR) reconstruction methods typically used for low-cost optical components. The experimental validations are performed with a geometrically complex target.

This paper presents a preliminary analysis of candle flame impact on the ultra-broadband terahertz (THz) communication links across a spectrum of interest from 300 GHz to 310 GHz. This approach is based on complex transfer functions extracted from channel measurements using a vector network analyzer (VNA) to study the variations in total received power and phase. The channel measurements are performed in a lecture room under line-of-sight (LoS) environment. By evaluating the results, it turns out that the ray trajectories remain unchanged with candle flame but have slightly altered amplitudes of paths and phases.

In this paper, we introduce a promising high resolution multistage approach for THz cameras for the detection and curvature extraction of metal fragments. The capability of this technique is demonstrated by detecting a 4 mm × 7 mm large metal fragment placed on a plank of beechwood.

In this paper, we propose a fast terahertz time-domain imaging method using a radar migration algorithm. We demonstrate high-resolution imaging in reflection without any collimating or focusing optics in the terahertz beam. In the proposed method, the sample is illuminated with a divergent terahertz beam, and the receiver collects both specular and diffuse reflections. We further present calibration and post-processing methods that allow us to compensate for the inherently low signal-to-noise ratio of an unfocused terahertz beam. The feasibility of the novel imaging method is demonstrated with geometrically complex samples and a fast terahertz time-domain spectroscopy system based on electronically controlled optical sampling. We show that our concept is capable of generating images of the objects regardless of their size, shape, orientation and position relative to the transmitter and receiver antennas. Objects with edge lengths well below 400 μm can be clearly detected. The method presented here thus lends itself to arbitrary scenarios and antenna configurations.

In this paper, we employ an ultrafast THz time-domain spectroscopy (TDS) system for computationally efficient imaging using a Kirchhoff migration algorithm. The imaging method makes use of uncollimated, unfocused terahertz beams, and eliminates the need for any opto-mechanical elements. Moreover, this method compensates for the low signal-to-noise ratio obtained with an uncollimated THz beam. We use a THz TDS system based on electronically controlled optical sampling (ECOPS), which achieves a measurement rate of 1600 traces per second. We validate the performance of the method with three different objects.

In this paper, a computationally efficient preprocessing approach for continuous wave THz spectroscopy based on Tukey windowing in the time-domain is proposed. Tukey windowing eliminates the effects of standing waves in the measurement setup while preserving the signal-to-noise ratio (SNR) and high frequency resolution of the spectroscopy system.

For the purpose of high-precision Radar Object Recognition (OR) system for real life emergency scenarios, a 60 GHz Super-Resolution FMCW radar imaging system is presented in this paper. Conventional radar imaging systems are limited by different hardware parameters such as bandwidth and antenna pattern resulting in distorted and noisy low-resolution (LR) images hindering the possibility of correct object recognition. Hence the radar imaging system proposed in this paper provides super-resolution (SR) images based on SR reconstruction methods typically used for low-cost optical components. Furthermore, the proposed SR radar system uses a low-cost single chip 60 GHz FMCW radar with two Rx antennas and one Tx antenna in a quasi monostatic configuration. The experimental validations are performed with geometrically complex targets by acquiring 3D radar images.

In this paper an ultra-wideband system is used to examine the effect of extinguishing water on a multilayer reflection concept. Additionally, the reflection measurements of typical indoor-materials with different water contents will be compared with the analytical results. The drawbacks of common characterization techniques, which require a multiple change in the measurement setup by sweeping the angle of incidence or supply just a narrow-band analysis are avoided. Furthermore, the reflection theory of multi-layer objects is briefly described so that the theoretical results are compared with real measurement. By the comparison of the measurement and the simulation a validation of this model is possible as well. The experimental validations are performed with two dielectric test objects. For the measurements the ZVB-20 network analyzer of Rohde & Schwarz is used for the analysis from 4.5 GHz to 13.5 GHz.

In this paper a novel estimation technique is introduced for a single measurement to estimate permittivity as well as the layer thickness of an object. The drawbacks of common characterization techniques, which require multiple changes in the measurement setup by sweeping the angle of incidence which just supply a narrow-band analysis of the permittivity are avoided. Furthermore, the reflection theory of multi-layer objects is briefly described so that the theoretical results are compared with measurements. By the comparison of the measurement and the simulation a validation of this model is possible. The introduced technique is ideal for autonomous security robotics where real-time conditions are necessary. Experimental validations are performed with two dielectric test objects. For the measurements the ZVB-20 network analyzer of Rohde & Schwarz is used working in a range from 4.5 GHz to 13.5 GHz.

For the purpose of Radar Object Recognition (OR) system for real life scenarios where only a partially reconstructed image of the Object Under Test is available, an OR Ultra-Wideband (UWB) Radar system is proposed. Conventional OR radar systems based on vector machines or neural networks result in a high recognition rates even for incomplete object images, but are not suitable for OR in a real time scenario due to the vast computational load. Hence the OR radar system proposed in this paper is based on a statistical approach employing a Bayesian detector and seven Object Recognition features with low mathematical and computational complexity. Furthermore, the proposed OR features are extracted from polarimetric radar images acquired by two imaging methods. Experimental validations are performed with an alphabet of twelve complex objects, a M-sequence UWB Radar device (4.5 GHz - 13.5 GHz) and two compact dual-polarized Ultra-Wideband antennas.

Daily newspapers all too often headline articles about millions in damages arising and several animals dying due to fires in barns and stables. Inadequate fire protection measures as well as a too late fire alarms are to blame for that. Automatic fire detection in such environments is a challenge especially due to the high dust load. As long as people are present in the stable, i.e. during the day, they can detect fires quickly. That is usually not the case during the night. In order to prevent this, it is important to design a system which can detect even the smallest amount of smoke. The most important point, however, is that this system shall not be triggered by dust because that would makes it useless for a stable-based application. In this paper a prototype of a smoke detector is presented which is suitable for as many different challenging environmental conditions as possible. This smoke detector incorporates different scattering properties of light at different aerosols, as described in [3].

This paper describes two new high selective optical approaches for the characterisation of critical aerosols in smoke detection applications, i.e. dust and water droplets. These approaches are based on the polarimetric analysis of different light scattering effects. The advantages of both methods are a high selectivity and low hardware costs. The first part of the paper gives an overview over the latest research into the polarimetric light scattering based methods for smoke detection applications at the Chair of Communication Systems of the University of Duisburg-Essen. The implementation of the described methods in a detector prototype will be described in a separate paper. The focus of the second part of the paper is the exploitation of the methods for the design of a test-procedure for the exposure of optical smoke detectors to mixed aerosols. Thus future smoke detectors could not only be tested in an environment with either smoke or dust, but also in a dusty scenario with an increasing concentration of smoke. This development keeps up with the latest advancements in detection technology.

In this paper a method for contour extraction of objects with a high degree of surface roughness in the range of 60 GHz Radar is presented. The proposed algorithm is based on an active contour model (snake) with external forces called component normalized gradient vector flow (CN-GVF). The snake algorithm extracts the contour of the object under test (OUT) based on wavefront Radar imaging methods. Experimental validations were performed with two fully integrated wideband frequency modulated continuous-wave (FMCW) single-chip Radar transceivers with an operational band from 57 GHz to 64 GHz, a corner cube retroreflector and a target object with a high degree of surface roughness.

In this paper a novel super-resolution wavefront extraction algorithm is introduced that merges the advantages of the efficiency correlation techniques based on [1] and super-resolution ability of DCM techniques [2]. The introduced ADCM algorithm is based on the correlation method DCM, but in contrast this algorithm uses a set of weighted differently shifted reference pulses. Due to the clustering of wavefronts the computational effort of the wavefront extraction is optimized. In the herein used set-up a low-cost, single chip radar operating from 57 GHz to 64 GHz is applied. The low-cost and low-weight radar is ideal for autonomous security robots where real-time conditions are necessary.

Material characterization utilizing microwave ellipsometry is based on the fact that the reflection coefficients of a wave impinging on a material depend on the incident fields polarization. Due to multiple reflections that occur in case of a material slab with finite dimensions these coefficients also strongly depend on the materials thickness. This paper describes an approach to determine the electromagnetic properties of a material under test by applying an ellipsometric measurement without knowledge of the materials thickness. The paper shows that the additional measurement of the transmission coefficients allows to perform an exact measurement of the complex permittivity since the thickness dependencies in reflection and transmission coefficients are removed. The measurements are done in the frequency range from 22 GHz to 26 GHz by using a vector network analyzer setup with a conical horn antenna.

The chaining of imaging techniques with material characterization capabilities is a very promising approach in the field of security applications. In this paper a polarimetric measurement setup which combines the advantages for imaging and material characterization is introduced. In the herein developed novel set-up the disadvantage of a bi-static configuration is overcome by the implementation of a retroreflector, allowing measurements applying a low-cost, single chip radar. This method is the basis for the combination of material characterization and imaging of targets with rough surface. Furthermore, a modified real time capable convergence algorithm based on [9] is proposed for the purpose of accurate imaging of rough objects. For the validation of this technique polarimetric measurements of real and rough objects are performed and the benefit of the polarimetric evaluation is shown.

In this paper a real time capable 2D imaging method for ultra-wideband Radar is presented. A well known wavefront localization method is adapted and improved using a super-resolution wavefront extraction method and exploring the polarimetric information of the target under test. The imaging algorithm is real time capable and directly maps an extracted wavefront to the target contour in contrast to classical popular migration algorithms. Furthermore, the new proposed imaging algorithm is designed for a circular scanning trajectory, or a rotating target and a bi-static antenna configuration with two receiver antennas. Experimental validations are performed with a geometrically complex object, a M-sequence UWB Radar device (4.5 GHz - 13.5 GHz) and compact Vivaldi UltraWideband antennas.

For the goal of an Object Recognition (OR) in emergency situations, an OR Ultra-Wideband (UWB) Radar system is proposed in this paper. Conventional OR Radar systems based on vector machines or neural networks result in a high recognition rates, but are not suitable for OR in a real time scenario, due to the vast computational load. Hence the OR Radar system proposed in this paper is based on a minimum mean square error detector and seven Object Recognition features with low mathematical and computational complexity. Furthermore, the proposed OR features are extracted from polarimetric images Radar acquired by two imaging methods. Experimental validations are performed with an alphabet of twelve complex objects, a M-sequence UWB Radar device (4.5 GHz - 13.5 GHz) and compact dual-polarized Ultra-Wideband antennas.