Inline image-based reinforcement detection for concrete additive manufacturing processes using a convolutional neural network

• DOI: 10.22260/isarc2024/0007
• Publication Date: 2024-05-27T13:41:51Z
• Authors: Lukas Johann Lach...; Lars Dittrich; Robin Dörrie; Harald Kloft; Annika Raatz; Tobias Recker

Inline image-based reinforcement detection for concrete additive manufacturing processes using a convolutional neural network Lukas Johann Lachmayer, Lars Dittrich, Robin Dörrie, Harald Kloft, Annika Raatz, Tobias Recker Pages 42-48 (2024 Proceedings of the 41st ISARC, Lille, France, ISBN 978-0-6458322-1-1, ISSN 2413-5844) Abstract: Within scope structural components, integration provides an inevitable opportunity to enhance load-bearing capacity elements. Besides rebar itself, ensuring as-planned cover is key achieving stable and long-term legally permissible integration. The thickness as-built however unpredictably altered during printing by varying material behaviour printed concrete. In addition, lack opportunities anchor elements before can lead displacement printing. this publication, we present approach determining position within additively manufactured components without any post-process measurement steps. During process, RGB images depth data are recorded camera mounted printhead. Subsequently, employed distinguish between structures deposited coloured image. By overlaying colour image with information, 3D point cloud generated, which marked. Keywords: Additive Manufacturing, Process Control, Image Processing, Neural Network, Printing Robot DOI: https://doi.org/10.22260/ISARC2024/0007 Download fulltext BibTex Endnote (RIS) TeX Import Mendeley