Spatter formation is a major issue in deep penetration welding with solid-state lasers at high welding speeds above 8 m/min. In order to limit spatter formation, the use of local gas flows represents a technically feasible solution. By using the gas flow, the pressure balance inside the keyhole, and therefore the keyhole stability, is affected. Existing investigations demonstrate a reduction in spatter and pore formation for partial penetration welding up to a welding speed of 5 m/min. However, the effect of the gas flow is not yet clarified for full penetration welding at welding speeds above 8 m/min. By using a precisely adjustable shielding gas supply, the effect of a local gas flow of argon was characterized by welding stainless steel AISI304 (1.4301/X5CrNi18-10). The influence of the gas flow on the melt pool dynamics and spatter formation was recorded by means of high-speed videography and subsequently analyzed by image processing. Schlieren videography was used to visualize the forming flow flied. By the use of the gas, a change in melt pool dynamics and gas flow conditions was observed, correlating to a reduction in loss of mass up to 70%. Based on the investigations, a model of the acting effect mechanism was given.

The projection of a point cloud onto a 2D camera image is relevant in the case of various image analysis and enhancement tasks, e.g., (i) in multimodal image processing for data fusion, (ii) in robotic applications and in scene analysis, and (iii) for deep neural networks to generate real datasets with ground truth. The challenges of the current single-shot projection methods, such as simple state-of-the-art projection, conventional, polygon, and deep learning-based upsampling methods or closed source SDK functions of low-cost depth cameras, have been identified. We developed a new way to project point clouds onto a dense, accurate 2D raster image, called Triangle-Mesh-Rasterization-Projection (TMRP). The only gaps that the 2D image still contains with our method are valid gaps that result from the physical limits of the capturing cameras. Dense accuracy is achieved by simultaneously using the 2D neighborhood information $(\mathit{rx},\mathit{ry})$ of the 3D coordinates in addition to the points $\mathit{P}(\mathit{X},\mathit{Y},\mathit{V})$. In this way, a fast triangulation interpolation can be performed. The interpolation weights are determined using sub-triangles. Compared to single-shot methods, our algorithm is able to solve the following challenges. This means that: (1) no false gaps or false neighborhoods are generated, (2) the density is XYZ independent, and (3) ambiguities are eliminated. Our TMRP method is also open source, freely available on GitHub, and can be applied to almost any sensor or modality. We also demonstrate the usefulness of our method with four use cases by using the KITTI-2012 dataset or sensors with different modalities. Our goal is to improve recognition tasks and processing optimization in the perception of transparent objects for robotic manufacturing processes.

Laser beam welding can produce narrow, high-quality welds in various industrial joining processes. The thermal expansion and contraction of the metal during the weld results in the displacement of the sheets. That leads to the formation of joint gaps and subsequent to a process interruption. This behavior has only been analyzed to a limited extent and causes manufacturers to rely on heavy clamping systems rather than using more flexible fixtureless approaches. Due to the time-consuming and costly nature of recording and producing erroneous weld seams, such recordings and datasets are rarely available in this area. This often limits the research towards adaptable fixtureless welding setups. Because of this, we present a multi-modal dataset consisting of 100 recorded welds that tracks the metal sheets movement. The developed setup enables the determination of boundary conditions for fixtureless welding. Two types of sensors record the welding process. First, three inductive probes are applied to record the metal sheets` movement and second, a long-wave infrared (LWIR) camera records changes in the thermal radiation field. Two different welding speeds and laser powers were used to produce a variety of welds. The dataset can be used for data-driven algorithms to predict the metal movement, analyze the thermal radiation field, or develop quality control methodologies.

Estimating depth from images is a common technique in 3D perception. However, dealing with non-Lambertian materials, e.g., transparent or specular, is still nowadays an open challenge. However, to overcome this challenge with deep stereo matching networks or monocular depth estimation, data sets with non-Lambertian objects are mandatory. Currently, only few real-world data sets are available. This is due to the high effort and time-consuming process of generating these data sets with ground truth. Currently, transparent objects must be prepared, e.g., painted or powdered, or an opaque twin of the non-Lambertian object is needed. This makes data acquisition very time consuming and elaborate. We present a new measurement principle for how to generate a real data set of transparent and specular surfaces without object preparation techniques, which greatly reduces the effort and time required for data collection. For this purpose, we use a thermal 3D sensor as a reference system, which allows the 3D detection of transparent and reflective surfaces without object preparation. In addition, we publish the first-ever real stereo data set, called TranSpec3D, where ground truth disparities without object preparation were generated using this measurement principle. The data set contains 110 objects and consists of 148 scenes, each taken in different lighting environments, which increases the size of the data set and creates different reflections on the surface. We also show the advantages and disadvantages of our measurement principle and data set compared to the Booster data set (generated with object preparation), as well as the current limitations of our novel method.

This paper proposes a FPGA-accelerated lens undistortion and rectification algorithm for a stereo-based phase measuring profilometry system. After a brief system overview we go into detail on the proposed hardware architecture for the stereo rectification block. The hardware architecture is based on a Xilinx Zynq-7020 SoC and uses compressed rectification maps to reduce the memory load. As a result, it is able to rectify a stereo setup with 435 mm baseline and converged optical axis of 32° between the cameras. The cameras are configured with 2 Mpix @ 60 fps and rectification can be applied "on the fly" during image grabbing.