16-Oct-2024 Image https://fordatis.fraunhofer.de/handle/fordatis/420
http://dx.doi.org/10.24406/fordatis/370

SYNOSIS dataset

SYNOSIS: Image synthesis pipeline for machine vision in metal surface inspection

Fulir, Juraj (Fraunhofer-Institut für Techno- und Wirtschaftsmathematik ITWM); Jeziorski, Natascha (Fraunhofer-Institut für Techno- und Wirtschaftsmathematik ITWM); Bosnar, Lovro  ; Hagen, Hans  ; Redenbach, Claudia (Fraunhofer-Institut für Techno- und Wirtschaftsmathematik ITWM); Gospodnetic, Petra (Fraunhofer-Institut für Techno- und Wirtschaftsmathematik ITWM); Herrfurth, Tobias  ; Trost, Marcus  ; Gischkat, Thomas  

ITWM Fraunhofer-Institut für Techno- und Wirtschaftsmathematik
Files in This Item:
FileDescriptionSizeFormat 
Dataset Teaser (samples).pdfTeaser images of all the textures and object instances.3,33 MBAdobe PDFPreviewDownload/Open
Real.zipReal part of the dataset.541,96 MBZIPDownload/Open
SynthPreprocess_CNR0_05.zipSynthetic part of the dataset.730,13 MBZIPDownload/Open
Readme_Real.mdReadme file on the usage of real part of the dataset.3,02 kBTextDownload/Open
Readme_Synth.mdReadme file on the usage of synthetic part of the dataset.3,16 kBTextDownload/Open

Abstract

Visual surface inspection images of aluminum test objects with different manufacturing surface textures and defects. The textures were produced by sandblasting, paralel milling and spiral milling. The dataset was used to support study of influence of surface texture on defect recognition. It contains real images, synthetic images and segmentation masks for both. Objects were inspected using a ring light and grayscale matrix camera monted on a robotic manipulator.

Technical Information

Dataset is a collection of PNG images and text files easily readable by any software.
Classification
005 Computerprogrammierung, Programme, Daten
671 Metallverarbeitung und Rohprodukte aus Metall
Keywords
Visual Surface Inspection
Synthetic Data
Surface Defect
Surface Texture
Metal Surface
Funder
Bundesministerium für Wirtschaft und Klimaschutz BMWK (Deutschland)
Fraunhofer Group
IUK-Technologie
Show full item record


Items in Fordatis are protected by copyright, with all rights reserved, unless otherwise indicated.