AI-BASED INSPECTION OF RECYCLED CARBON FIBRE FABRIC
Authors/Creators
Description
The dataset is part of an EU-funded project and is included in the paper entitled "AI-based Inspection of Recycled Carbon Fibre Fabric."
The work presented in this paper is related to the project “MC4” and has received funding from the European Union’s Horizon Europe research and innovation program under grant agreement No 101057394. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union. Neither the European Union nor the granting authority can be held responsible for them.
sample_images.zip - It is a zip file containg the .png images. These are the samples images of the output from the in-house developed sensor.
dataset_samples.pkl - It is a pickle file. This contains the dataset whose results are reported in the paper .
load_data.py - This is the sample code to load the data. Explanation to the shapes and the structure of the data are present here.
Abstract (English)
Carbon fiber fabric for structural components has to satisfy high quality requirements. The automation of end-of-line quality control is thus an important step towards the automated documentation and correction of defects.
In order to scan fabric already during the production process, a modular vision-based sensor is presented. The sensor generates multi-modal images, including reflectance properties and fiber orientation. The sensor concept can be extended up to 2.5m in length (100”) to allow the scanning of the whole width of the fabric. An AI-based data analysis is developed and assessed on a range of typical defects that occur during the manufacturing of carbon fiber fabric. The paper describes the approach to the sensor design, the image modalities and the AI-based method for defect segmentation and classification.
A specific analysis is also done in relation to the properties of woven and nonwoven materials made from recycled carbon fiber. These materials have higher visual variability and the analysis needs to take this into account.
Files
sample_images.zip
Additional details
Dates
- Created
-
2024-05-15