Natalia Pérez García-De-La-Puente

Mid-infrared Optical Coherence Tomography (MIR-OCT) is a promising Non-Destructive Testing (NDT) technique due to its high-resolution imaging capabilities and extensive applicability across various industrial domains. Studies developing Deep Learning (DL) models to detect defects in MIR-OCT scans are scarce, and few have been used for ceramic quality assessment. To address this gap, we introduce the MIR-OCT Scan Dataset for Ceramic Quality Assessment (CeraMIRScan), including labels to detect and segment defects. The dataset comprises 29 volumes corresponding to MIR-OCT scans of 3D printed ceramic pieces, decomposed into 21,882 B-scan images, each paired with expert-annotated binary masks capturing defects such as pores, delaminations, and inclusions. Notably, 41.38% of the images contain visible anomalies. To illustrate the dataset's applicability to DL, we provide baseline segmentation results using a U-Net architecture, achieving 80.55% precision, 80.00% recall, and an 80.27% Dice score. CeraMIRScan offers a benchmark resource for advancing automated quality assessment and supports the development of MIR-OCT-based defect characterization methods.

Non-destructive testing (NDT) is essential in ceramic manufacturing to ensure the quality of components without compromising their integrity. In this context, Optical Coherence Tomography (OCT) enables high-resolution internal imaging, revealing defects such as pores, delaminations, or inclusions. This paper presents an automatic defect detection system based on Deep Learning (DL), trained on OCT images with manually segmented annotations. A neural network based on the U-Net architecture is developed, evaluating multiple experimental configurations to enhance its performance. Post-processing techniques enable both quantitative and qualitative evaluation of the predictions. The system shows an accurate behavior of 0.979 Dice Score, outperforming comparable studies. The inference time of 18.98 seconds per volume supports its viability for detecting inclusions, enabling more efficient, reliable, and automated quality control.

This paper aims to evaluate mid-infrared (MIR) Optical Coherence Tomography (OCT) systems as a tool to penetrate different materials and detect sub-surface irregularities. This is useful for monitoring production processes, allowing Non-Destructive Inspection Techniques of great value to the industry. In this exploratory study, several acquisitions are made on composite and ceramics to know the capabilities of the system. In addition, it is assessed which preprocessing and AI-enhanced vision algorithms can be anomaly-detection methodologies capable of detecting abnormal zones in the analyzed objects. Limitations and criteria for the selection of optimal parameters will be discussed, as well as strengths and weaknesses will be highlighted.

This paper presents the investigation and results of the training and use of the popular object detectors YOLOv8 and YOLOv11 in the identification of anomalies during the painting process of wind turbines. Several data augmentation techniques have been performed on the original dataset to assess its impact during the training process and the evaluating metrics of the proposed object detectors. The classification process has been carried out through binary scenarios to determine how the data augmentation process can enhance the use of neural network behavior in real-time applications. The aim of this paper is to quantify and evaluate how the introduction of new data augmentation enhances the performance of existing state-of-the-art models relying on data augmentation and labelling techniques.