Search

Light detection and ranging (LiDAR) technology plays an important role in achieving higher levels of autonomous driving. These sensors, although robust in clear weather conditions, including night scenes, tend to degrade in adverse weather conditions like fog, rain and snowfall. An autonomous vehicle relying on LiDAR should be able to assess in a real-time manner its limitations and raise an alarm in such scenarios. In this paper, we present a comprehensive statistical data analysis of the effects of real-world adverse weather conditions on the properties of LiDAR point clouds. Namely, we analyze the effect on range, reflectance and resolution of objects in point clouds recorded by LiDAR in challenging weather conditions. Furthermore, based on the results of the analysis, we propose RangeWeatherNet, a lightweight deep convolutional network architecture for classification of weather and road conditions. The classification accuracy of our network outperforms the existing methods by a large margin (+11.8 %). The core network runs at 102 fps, which with the data preprocessing step, amounts to total 32 fps, which is higher than the usual LiDAR acquisition rate. To the best of our knowledge, this is the first approach that uses deep learning for classification of weather conditions on LiDAR point clouds.

Flash lamp annealing (FLA) is a modern annealing technique which, starting from microelectronics, has spread over new application areas like flexible electronics, photovoltaics or thin film deposition. Because of the short annealing time in the range of milliseconds and below, FLA allows the suppression of unwanted processes like diffusion, the annealing of temperature-sensible substrates, and the saving of process time and energy. In addition, it is predestined for roll-to-roll applications. However, the determination of the thermal cycle during FLA is challenging. The existing methods for a direct temperature measurement, mostly based on pyrometry, are elaborate and have to solve the problem to detect thermal radiation against the background of the intense flash light. An alternative way is simulation, but now an extended knowledge about the flash and the material system to be flashed is needed. In this work we describe a methodology to determine the thermal cycle during FLA without the need for a direct temperature measurement. This methodology is based on an optical-thermodynamic simulation and calibration experiments which can be implemented with reasonable effort under certain assumptions. The simulation considers not only the properties of the flash and the sample, but also the reflectivity of the chamber walls. Finally, the pros and cons of this methodology are shortly discussed.

Drug reaction with eosinophilia and systemic symptoms (DRESS) is an infrequent, but severe, adverse drug-induced reaction which occurs due to massive T-cell stimulation resulting in cytotoxicity and eosinophil activation and recruitment. The incidence is 0.4 cases per 100, 0000 in the general population; the mortality rate is up to 10%. Therefore, we believe that recognizing this syndrome is of particular importance. The problem we notice is that DRESS is often seen and described in patients receiving rheumatologic or anticonvulsant drugs, but very rarely in psychiatric hospitals, where Clozapine is frequently used, and that is the importance of this paper. DRESS Syndrome must be recognized promptly, and causative drugs withdrawn. Indeed, it has been reported that the earlier the drug withdrawal, the better the prognosis. In this paper, we present three cases of Clozapine-induced DRESS. All cases were recorded in the Multicenter Drug Safety Surveillance Project (AMSP).

Thermal processing in the millisecond range provides advanced, non-equilibrium annealing techniques which allow dedicated material modifications at the surface without affecting the substrate volume below. The process called flash lamp annealing (FLA) is one of the most diverse methods of short time annealing with applications ranging from the classical field of semiconductor doping to the treatment of layers on glass, polymers and other flexible substrates. It still continues to extend to other material classes and applications, and becomes of interest for an increasing number of users. Other phrases for FLA used throughout the literature are intense pulsed light sintering (IPL) or photonic curing. This review presents a short and comprehensive view of the current state of the art of FLA with a focus on functional coatings. After an introduction including historical aspects a look is taken to equipment issues as well as to the pioneering role which semiconductor processing in the framework of advanced chip technology played for the development of short time annealing. Mostly, examples of processing for photovoltaics, including doping aspects, hydrogen engineering, copper indium gallium diselenide (CIGS), silicon crystallization on glass, and transparent conductive oxides (TCO), including indium tin oxide (ITO), zinc oxide (also Al-doped AZO) as well as inkjet printing for flexible electronics will be presented.

Developing vehicle collision warning systems on mobile devices aiming to alert drivers about driving environments, and possible collision with other vehicles has become more and more popular. In these systems, robust and reliable vehicle detection is a critical step. This paper presents a vision-based vehicle detection system using a deep learning approach on mobile platforms. Our focus is on the mobile system with camera which is mounted on the vehicle. Integrating detection with tracking is also discussed to illustrate the benefits of deep learning for vehicle detection. Finally, we present the high efficient experimental results based on mobile device mounted on a car.

The flexible and stretchable electronic market is increasing particularly in the field of biomedical electronics. Widely used printed silver conductive tracks today are only eligible for on-skin applications. However, for biomedical applications fully biocompatible, flexible and even stretchable materials for device fabrication are needed. This paper presents an additive printing approach to fabricate flexible and stretchable electronics by using a biocompatible platinum material. Usually, in order to realize electrically conducting Pt-interconnects by inkjet printing, it requires a furnace sintering at prohibitively high temperatures, which are not compatible with thermal sensitive polymeric substrates. This paper describes a high-power diode laser sintering (HPDL) and a flash lamp annealing (FLA) as promising alternative sintering methods. Both processes are eligible whereas laser sintering showed slightly better results. Bending tests and adhesive strength tests of platinum printed inks on polyimide with up to 180 000 cycles, show that printed platinum is a suitable biocompatible material for flexible electronics.

Water reservoirs around the world suffer from accelerated sediment loads and, consequently, contamination. Notably, in water-scarce regions such as Jordan, this poses a threat to an important water source, and identifying the sediment sources is an important task. Thus, a sediment fingerprinting study in the Wadi Al-Arab catchment of northern Jordan was implemented with special attention directed to the development of suitable correction factors necessary to improve the comparability of source and sink sediments. The selection of seven conservative elements for the sediment fingerprinting was made, with specific attention directed to the chemical environment of the reservoir. Thirty-six samples from six different surface and subsurface sources and 38 sink samples from the Wadi Al-Arab reservoir were collected. In total, 27 organic and inorganic elements as well as radionuclides were analysed. Two vertical physicochemical water profiles provided information on the pH and Eh conditions and common element concentrations. The stepwise multiple regression analysis model (SMRAM) was developed to explore parameters that influence the element concentrations and their interrelations, and to calculate an element-specific correction factor. The standard selection procedure was expanded by the comparison of water and sink sediment element concentrations, a literature review concerning the pH and Eh conditions and, in selected cases, a correlation analysis. The combination of Al, Cr, Mn, Fe, 232Th, 228Th and 137Cs provided the best source discrimination, and based on Monte Carlo simulations, the mixing model revealed the existence of three major sediment source areas. These are as follows: (i) olive orchards on slopes, which delivered 59 ± 8 % of the sediments in the sink; (ii) cultivated fields on plateau and saddle positions contributed 11 ± 9 %; and (iii) slopes with natural vegetation used for grazing contributed 29 ± 15 % of the deposited sediment. With a mean residual error of 1.04 %, the sum of the source concentrations differs only slightly from sink concentrations and proves that the model is reliable. The SMRAM model revealed that the different inorganic (total inorganic carbon, TIC) and organic (total organic carbon, TOC) carbon contents and the clay/sand content influence the element concentrations of the sediment samples. Due to the carbonatic environment, it was mainly necessary to correct for TIC. Applying an expanded literature review regarding the chemical environment under investigation, in addition to the standard mass conservation and Kruskal-Wallis test, prevented possible non-conservative elements from entering the discriminant analysis.

This paper proposed a monocular vehicle detection for forward collision warning system. We use the active-learning framework to train a cascade classifier and use a two steps vehicle detection. We used five test data to quantify our detection performance, analyzing the two-stage vehicle detection improvement, and the overall detection rate and the false detection rate. In a good light condition, the detection rate and the false detection rate can achieve 0.967 and 0.122, respectively. Our system can achieve up to 45 frames per second on Intel core I7-6700 CPU.

Today automotive headlights are using high current Light Emitting Diodes (LEDs). LED drivers are necessary to provide the appropriate voltage out of the car battery. Recent microcontroller-based solutions are costly and power hungry. This paper presents a power-efficient LED driver. The design is CISPR25 class-3 EMI compliant and supports ISO 7637-2 and ISO 16750-2 standards. In addition, the system includes a special feature of Dynamic Headroom Control (DHC) that aims at maximizing overall system efficiency by adjusting voltage at the boost controller by sensing the LEDs voltages. Experimental results show up to 88% power-efficiency.

Logarithmic spirals are found on different length scales in nature, e.g., in nautilus shells, cyclones, and galaxies. The underlying formation laws can be related to different growth mechanisms, pressure gradients, and density waves. Here, we report on the self-organized formation of symmetric logarithmic crystallization spirals in a solid material on the micrometer length scale, namely, in an amorphous Ge:Mn layer on a Ge substrate. After exposure to a single light pulse of a flashlamp array, the Ge:Mn layer is crystallized and reveals a partially rippled surface and logarithmic microspirals. Finally, we present a model describing the formation of the crystallization spirals by directional explosive crystallization of the amorphous Ge:Mn layer, which is triggered by the flashlamp light pulse.

We present a study of the tunnel magneto-Seebeck (TMS) effect in MgO based magnetic tunnel junctions (MTJs). The electrodes consist of CoFeB with in-plane magnetic anisotropy. The temperature gradients which generate a voltage across the MTJs layer stack are created using laser heating. Using this method, the temperature can be controlled on the micrometer length scale: here, we investigate, how both, the TMS voltage and the TMS effect, depend on the size, position and intensity of the applied laser spot. For this study, a large variety of different temperature distributions was created across the junction. We recorded two-dimensional maps of voltages generated by heating in dependence of the laser spot position and the corresponding calculated TMS values. The voltages change in value and sign, from large positive values when heating the MTJ directly in the centre to small values when heating the junction on the edges and even small negative values when heating the sample away from the junction. Those zero crossings lead to very high calculated TMS ratios. Our systematic analysis shows, that the distribution of the temperature gradient is essential, to achieve high voltage signals and reasonable resulting TMS ratios. Furthermore, artefacts on the edges produce misleading results, but also open up further possibilities of more complex heating scenarios for spincaloritronics in spintronic devices., 12 pages, 6 figures

Three-dimensional topological insulators are fascinating materials with insulating bulk yet metallic surfaces that host highly mobile charge carriers with locked spin and momentum. Remarkably, surface currents with tunable direction and magnitude can be launched with tailored light beams. To better understand the underlying mechanisms, the current dynamics need to be resolved on the timescale of elementary scattering events (∼10 fs). Here, we excite and measure photocurrents in the model topological insulator Bi2Se3 with a time resolution of 20 fs by sampling the concomitantly emitted broadband terahertz (THz) electromagnetic field from 0.3 to 40 THz. Strikingly, the surface current response is dominated by an ultrafast charge transfer along the Se–Bi bonds. In contrast, photon-helicity-dependent photocurrents are found to be orders of magnitude smaller than expected from generation scenarios based on asymmetric depopulation of the Dirac cone. Our findings are of direct relevance for broadband optoelectronic devices based on topological-insulator surface currents., Surface currents in topological insulators can be controlled by light, but the underlying mechanisms are not well understood. Here, Braun et al. report an ultrafast shift photocurrent at the surface of Ca-doped Bi2Se3, whereas injection currents are much smaller than expected from asymmetric depopulation of the Dirac cone.

3D vision is becoming increasingly important, in major motion pictures as well as in home entertainment. Depth map generation plays a major role in 2D-to-3D conversion and acts as a precursory stage to Depth Based Image Rendering (DIBR). This paper presents a novel algorithm for depth map generation using a block based approach. The algorithm is composed of 2 parts: Background depth map generation and moving object detection & filling. The proposed algorithm has been implemented both in software as well as in VLSI.

In this paper, we propose an improved dynamic gamma correction method to enhance contrast and image quality for Liquid Crystal Displays (LCD) with Multi-Phosphor White Light Emitting Diodes (MPW LED). In order to reduce the hardware cost and improve the contrast and the edge strength, the proposed gamma correction method uses content-based analysis to reduce computational complexity. Experimental results demonstrate that the proposed gamma correction method has better visual assessment. Absolute Mean Brightness Error (AMBE), contrast enhancement, and similar Edge Loss Rate (ELR) quality are superior to the traditional methods.

Stereo vision has been used in many different applications such as robot, entertainment, and car electronics. Disparity-based distance estimation requires a large amount of computational resources because of high block-matching computations, especially on the high-resolution image pair. In this paper, we propose an algorithm for computing the distance between our own driving car and the front car for vehicle collision avoidance applications. Our system is designed to focus on the collision avoidance with the front end object which includes horizontal displacement. Therefore, the target of the front end of system is to find out the position and movement of these objects.

Shallow, Boron (B)-doped p+ emitters have been realized using spin-on deposition and Flash Lamp Annealing (FLA) to diffuse B into monocrystalline float zone Silicon (Si). The emitters extend between 50 and 140 nm in depth below the surface, have peak concentrations between 9x1019 cm-3 and 3x1020 cm-3, and exhibit sheet resistances between 70 and 3000 Ohm/Square. An exceptionally large increase in B diffusion occurs for FLA energy densities exceeding approximately 93 J/cm2 irrespective of 10 or 20 ms pulse duration. The effect is attributed to enhanced diffusion of B caused by Si interstitial injection following a thermally activated reaction between the spin-on diffusant film and the silicon wafer.