News

Presentation on digital photogrammetry and laboratory tour for the geomatics team at the workshop "Environmental Risk Assessment and Mitigation on Cultural Heritage Assets in Central Asia" (ERAMCA) in Weimar.

On the occasion of the Open Lab Night 2022 the chair of Computer Vision opens its doors to the public for the first time to the 3D Reality Capture Scan Lab (3ReCapSL).
Equipped with carefully orchestrated cutting-edge imaging equipment such as high-resolution digital system cameras and high-speed video sensors, 3ReCapSL offers a glimpse into the present and the future of image-based 3D reconstruction of static and dynamic scenes, motion and deformation fields. During the Open Lab Night 2022, researchers of the Computer Vision chair demonstrate live applications of state-of-the-art techniques in the field of accurate and photorealistic 3D reconstruction. Get an impression of current research and seize the opportunity to capture a three-dimensional selfie with more than 100 cameras.

The relevance of moving platforms for image acquisition - such as mobile phones or drones - is steadily increasing. The practical image resolution achieved by a moving camera is, however, typically worse than the theoretically possible resolution. Effects of motion blur, out of focus, noise, and improper camera parameters can distinctly reduce image sharpness and, thereby, the capability of representing fine structures. In this project, the practical resolution will be estimated by means of a Siemens star and related to camera movement.

Modern 3D reconstruction algorithms opened up a wide range of applications to capture real life geometry. Machine learning based methods are used to segment and understand these reconstructed 3D scenes. To train such methods, a vast amount of cleaned, annotated and curated data is needed, which is a tedious and time-consuming task. In this semester's project, students demonstrate methods to synthesize a large amount of 3D indoor scenes using freely available CAD models and extract suitable representations of point clouds, viewpoints and textual scene descriptions.

Julian Steinhaußen will give a presentation on "Deep Learning for Image Registration" on July 12, 2022 at 2pm as part of his master's defense.

7th "Schule MIT Wissenschaft Thüringen" conference from June 30 to July 2 at the Leonardo Hotel Weimar for Thuringian teachers with drone demonstration and joint lecture on testing aging infrastructure structures with digital technologies.

For Engineering Day, the geodesy lecture will be moved to LH A of M13c on June 29, 2022 at 7:30 am.

Presentation of interim results in the ev.AI.luate project at the D-A-CH Joint Meeting for Call 2021 - Bridges and Soils.

As part of the 3rd Thuringian Digital Festival in Jena, the "Digital and Open Source Prize Thuringia 2022" was awarded. Thuringia's Minister of Economic Affairs, Wolfgang Tiefensee, awarded the €5,000 prize in the "Digital Products & Services" category to Infralytica GmbH from Weimar.

Christian Benz will present his work on "Image-based Control Point Detection and Correspondence-free Georeferencing" on June 10, 2022 at 9:15 am at the XXIV ISPRS Congress in Nice, France.