Fraunhofer-Institut für Optronik, Systemtechnik und Bildauswertung IOSB

Filters

4
Reset filters

Settings
Now showing 1 - 10 of 4369
  • Publication
    Optisch nichtlineare Materialien für Anwendungen im Bereich Laserschutz - eine detaillierte Analyse der optischen Leistungsbegrenzung
    (KIT, 2024-06-14)
    Dengler, Stefanie
    Durch eine immer weitere Verbreitung von Lasern in nimmt auch die Notwendigkeit, den Menschen und sensitive Geräte und Optiken vor Laserstrahlung effizient zu schützen, ständig zu. Ein mögliches Schutzkonzept kann durch die Verwendung optisch nichtlinearer Materialien realisiert werden. In der vorliegenden Arbeit werden verschiedene optisch nichtlineare Materialien aus den Materialgruppen kohlenstoffbasierter Nanomaterialien, organometallischer Moleküle und metallischer Nanopartikel untersucht.
  • Publication
    Comparative analysis of C²ₙ estimation methods for sonic anemometer data
    ( 2024-06-01)
    Beason, M.
    ;
    Potvin, G.
    ;
    Sprung, Detlev
    ;
    McCrae, J.
    ;
    Gladysz, Szymon
    Wind speed and sonic temperature measured with ultrasonic anemometers are often utilized to estimate the refractive index structure parameter C²ₙ, a vital parameter for optical propagation. In this work, we compare four methods to estimate C²ₙ from CT2 , using the same temporal sonic temperature data streams for two separated sonic anemometers on a homogenous path. Values of C²ₙ obtained with these four methods using field trial data are compared to those from a commercial scintillometer and from the differential image motion method using a grid of light sources positioned at the end of a common path. In addition to the comparison between the methods, we also consider appropriate error bars for C²ₙ based on sonic temperature considering only the errors from having a finite number of turbulent samples. The Bayesian and power spectral methods were found to give adequate estimates for strong turbulence levels but consistently overestimated the C²ₙ for weak turbulence. The nearest neighbors and structure function methods performed well under all turbulence strengths tested.
  • Publication
    Optimization of the slope efficiency of a core-pumped thulium-doped fiber laser by the thermally diffused expanded-core technique
    ( 2024-05-15)
    Panitzek, Dieter
    ;
    Romano, Clément
    ;
    Eichhorn, Marc
    ;
    Kieleck, Christelle
    We report on the mode field adaption of an active thulium-doped fiber by using the thermally-diffused expanded-core technique. The fiber core diffusion is analyzed by splice transmission measurements and visually from side view images. The obtained heating parameters are used to build a thulium-doped fiber laser emitting at 2036nm that is core-pumped by an erbium:ytterbium fiber laser. By allowing the fiber cores to diffuse, the mode fields of the active and passive fibers are adapted for both the signal and pump wavelength. The adaptation of the mode fields increases the slope efficiency from 66.1% to 75.0%. The obtained slope efficiency is close to the stoke efficiency of 77.0%. By comparing the results with a fiber laser simulation, the slope efficiency of 75.0% is verified to be the maximum slope efficiency taking the active fiber length into account.
  • Publication
    Simulation of the reflection of a high energy laser beam at the sea surface for hazard and risk analyses
    ( 2024-04-10)
    Schwenger, Frederic
    ;
    Azarian, Adrian
    ;
    Henrichsen, Michael
    The application of a high energy laser beam in a maritime scenario necessitates a laser safety concept to prevent injury to personnel or uninvolved third parties from uncontrolled reflections of laser light from the sea surface. Therefore, it is crucial to have knowledge of the amount and direction of reflected laser energy, which varies statistically and depends largely on the dynamics of the wavy sea surface. These dynamics are primarily influenced by wind speed, wind direction, and fetch. An analytical model is presented for calculating the time-averaged spatial intensity distribution of the laser beam reflected at the dynamic sea surface. The model also identifies the hazard areas inside which laser intensities exceed a fixed exposure limit. Furthermore, as far as we know, our model is unique in its ability to calculate the probabilities of potentially eye-damaging glints for arbitrary observer positions, taking into account the slope statistics of gravity waves. This is a critical first step toward an extensive risk analysis. The simulation results are presented on a hemisphere of observer positions with fixed radii from the laser spot center. The advantage of the analytical model over our numeric (dynamic) model is its fast computation time. A comparison of the results of our new analytical model with those of the previous numerical model is presented.
  • Patent
    Network node for a non-detectable laser communication system
    ( 2024-02-07)
    Kieleck, Christelle
    ;
    Eichhorn, Marc
    ;
    Rudow, Oliver
    ;
    Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V.
    ;
    Hensoldt Sensors GmbH
    A network node (120) for a non-detectable laser communication system (100), wherein the laser communication system (100) is configured to send to the network node (120) at least one laser beam (10), comprises a reflector device (123), configured to generate, by a reflection of the laser beam (10), a reflected laser beam (20), and a modulator device (125), configured to provide a modulation of the reflected laser beam (20).
  • Publication
  • Publication
    Sensor-based characterization of construction and demolition waste at high occupancy densities using synthetic training data and deep learning
    ( 2024)
    Kronenwett, Felix
    ;
    Maier, Georg
    ;
    Leiss, Norbert
    ;
    Gruna, Robin
    ;
    Thome, Volker
    ;
    Längle, Thomas
    Sensor-based monitoring of construction and demolition waste (CDW) streams plays an important role in recycling (RC). Extracted knowledge about the composition of a material stream helps identifying RC paths, optimizing processing plants and form the basis for sorting. To enable economical use, it is necessary to ensure robust detection of individual objects even with high material throughput. Conventional algorithms struggle with resulting high occupancy densities and object overlap, making deep learning object detection methods more promising. In this study, different deep learning architectures for object detection (Region-based CNN/Region-based Convolutional Neural Network (Faster R-CNN), You only look once (YOLOv3), Single Shot MultiBox Detector (SSD)) are investigated with respect to their suitability for CDW characterization. A mixture of brick and sand-lime brick is considered as an exemplary waste stream. Particular attention is paid to detection performance with increasing occupancy density and particle overlap. A method for the generation of synthetic training images is presented, which avoids time-consuming manual labelling. By testing the models trained on synthetic data on real images, the success of the method is demonstrated. Requirements for synthetic training data composition, potential improvements and simplifications of different architecture approaches are discussed based on the characteristic of the detection task. In addition, the required inference time of the presented models is investigated to ensure their suitability for use under real-time conditions.
  • Publication
    A Cross Branch Fusion-Based Contrastive Learning Framework for Point Cloud Self-supervised Learning
    ( 2024)
    Wu, Chengzhi
    ;
    Huang, Qianliang
    ;
    Jin, Xingkun
    ;
    Pfrommer, Julius
    ;
    Beyerer, Jürgen
    Contrastive learning is an essential method in self-supervised learning. It primarily employs a multi-branch strategy to compare latent representations obtained from different branches and train the encoder. In the case of multi-modal input, diverse modalities of the same object are fed into distinct branches. When using single-modal data, the same input undergoes various augmentations before being fed into different branches. However, all existing contrastive learning frameworks have so far only performed contrastive operations on the learned features at the final loss end, with no information exchange between different branches prior to this stage. In this paper, for point cloud unsupervised learning without the use of extra training data, we propose a Contrastive Cross-branch Attention-based framework for Point cloud data (termed PoCCA), to learn rich 3D point cloud representations. By introducing sub-branches, PoCCA allows information exchange between different branches before the loss end. Experimental results demonstrate that in the case of using no extra training data, the representations learned with our self-supervised model achieve state-of-the-art performances when used for downstream tasks on point clouds.
  • Publication
  • Publication
    Utilizing Dataset Affinity Prediction in Object Detection to Assess Training Data
    ( 2024)
    Becker, Stefan
    ;
    Bayer, Jens
    ;
    Hug, Ronny
    ;
    Hübner, Wolfgang
    ;
    Arens, Michael
    Data pooling offers various advantages, such as increasing the sample size, improving generalization, reducing sampling bias, and addressing data sparsity and quality, but it is not straightforward and may even be counterproductive. Assessing the effectiveness of pooling datasets in a principled manner is challenging due to the difficulty in estimating the overall information content of individual datasets. Towards this end, we propose incorporating a data source prediction module into standard object detection pipelines. The module runs with minimal overhead during inference time, providing additional information about the data source assigned to individual detections. We show the benefits of the so-called dataset affinity score by automatically selecting samples from a heterogeneous pool of vehicle datasets. The results show that object detectors can be trained on a significantly sparser set of training samples without losing detection accuracy. T