Details

Autor(en) / Beteiligte

• Pandiyan, Vigneashwara
• Cui, Di
• Richter, Roland Axel
• Leparoux, Marc

Titel

Real-time monitoring and quality assurance for laser-based directed energy deposition: integrating co-axial imaging and self-supervised deep learning framework

Ort / Verlag

Zenodo

Erscheinungsjahr

2023

Link zum Volltext

Quelle

Alma/SFX Local Collection

Beschreibungen/Notizen

• The experimental setup utilized a co-axial color Charged Couple Device (CCD) camera, integrated into the laser deposition head. This camera operates at a frame rate of 30 frames per second and captures the morphology of the process area. The captured images consist of three RGB channels with a 640 × 480 pixels resolution. To enable the camera to capture the radiation from the process zone, a beam splitter is installed on Precitec's laser applicator head. An optical notch filter within the 650–675 nm range also blocks the laser wavelengths. The dataset consists of four categories that covers the process map of DED process [.rar file].The dataset consist of around 48,000 images that are labelled into 4 categories [P1-P2-P3-P4]. The images correspond to DED process zone captured co-axiallyThe categories are function of linear laser energy deposited. The folder is already split into Train and Test.