cs.CVMar 11, 2026

Towards Cognitive Defect Analysis in Active Infrared Thermography with Vision-Text Cues

Authors: Mohammed Salah, Eman Ouda, Giuseppe Dell'Avvocato, Fabrizio Sarasini, Ester D'Accardi, Jorge Dias, Davor Svetinovic, Stefano Sfarra, 

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Active infrared thermography (AIRT) is currently witnessing a surge of artificial intelligence (AI) methodologies being deployed for automated subsurface defect analysis of high performance carbon fiber-reinforced polymers (CFRP). Deploying AI-based AIRT methodologies for inspecting CFRPs requires the creation of time consuming and expensive datasets of CFRP inspection sequences to train neural networks. To address this challenge, this work introduces a novel language-guided framework for cognitive defect analysis in CFRPs using AIRT and vision-language models (VLMs). Unlike conventional learning-based approaches, the proposed framework does not require developing training datasets for extensive training of defect detectors, instead it relies solely on pretrained multimodal VLM encoders coupled with a lightweight adapter to enable generative zero-shot understanding and localization of subsurface defects. By leveraging pretrained multimodal encoders, the proposed system enables generative zero-shot understanding of thermographic patterns and automatic detection of subsurface defects. Given the domain gap between thermographic data and natural images used to train VLMs, an AIRT-VLM Adapter is proposed to enhance the visibility of defects while aligning the thermographic domain with the learned representations of VLMs. The proposed framework is validated using three representative VLMs; specifically, GroundingDINO, Qwen-VL-Chat, and CogVLM. Validation is performed on 25 CFRP inspection sequences with impacts introduced at different energy levels, reflecting realistic defects encountered in industrial scenarios. Experimental results demonstrate that the AIRT-VLM adapter achieves signal-to-noise ratio (SNR) gains exceeding 10 dB compared with conventional thermographic dimensionality-reduction methods, while enabling zero-shot defect detection with intersection-over-union values reaching 70%.

Figures & tables

Fig. 1: Front-side view of the impacted specimens, subjected to low-velocity impact at 5 J and 15 J.
Fig. 2: Top: Front-side heating with pulsed flash lamps (A.1-A.2) and front-side long-pulse halogen heating (A.3). Bottom: Schematic representation of inspection setup.
Fig. 3: Top: Backside heating with flash (B.1) and halogen lamp (B.2). Bottom: Schematic representation of transmission inspection setup.
Fig. 4: Overview of the defect analysis framework. The inspection sequence is preprocessed to generate a compact single image representation similar to the VLMs pretraining domain. Consequently, a VLM generates bounding boxes for defects.
Defect ClassTemp.SamplesMetricRawTSRPCADAT1D-DCAE-AIRTC-AETOurs
5 JAmbient6Contrast0.2070.2410.3020.3660.3830.3610.478
SNR (dB)21.7524.5030.7133.4835.8334.1142.18
−70∘-70^{\circ} C7Contrast0.1980.2290.2890.3510.3660.3440.456
SNR (dB)20.9023.6229.4032.0534.2132.8840.27
15 JAmbient7Contrast0.2270.2870.3870.3950.4360.3870.534
SNR (dB)22.1124.7432.9538.2938.3736.5843.19
TABLE I: Quantified contrast and SNR for the AIRT-VLM adapter representation benchmarked against state-of-the-art thermography dimensionality reduction methods, TSR, PCA, DAT [ 22 ] , 1D-DCAE-AIRT [ 23 ] , and C-AET [ 29 ] , under ambient and low-temperature ( −70∘-70^{\circ} C) conditions.
TABLE II: Qualitative comparisons between state-of-the-art AIRT dimensionality reduction techniques: TSR [ 33 ] , PCA [ 34 ] , DAT [ 22 ] , 1D-DCAE-AIRT [ 23 ] , C-AET [ 29 ] , and the proposed AIRT-VLM adapter representation, under ambient and low-temperature ( −70∘-70^{\circ} C) conditions.
Fig. 5: Aggregate a) contrast and b) SNR on all 25 CFRP inspection sequences.
Fig. 6: IoU and normalized center distance achieved by Qwen-VL, CogVLM, and GroundingDINO when coupled with AIRT-VLM adapter on all 25 CFRP inspection sequences. The performances of these models are also presented when coupled with state-of-the-art AIRT dimensionality reduction methods.
TABLE III: Defect detection results of Qwen-VL, CogVLM, and GroundingDINO coupled with the AIRT-VLM adapter, under ambient and low-temperature. The results show consistent defect detection performances across all samples.
MetricAverage PoolingMax PoolingPCA
Contrast0.5220.4710.547
SNR (dB)42.8739.1842.41
IoU0.6910.5390.701
Normalized Center Distance0.01380.03780.0118
TABLE IV: Quantified performances of average, max, and PCA pooling operations.
MetricAverage PoolingNMS
IoU0.6910.707
Normalized Center Distance0.01380.0136
Execution Time (s)4.337.8
TABLE V: Defect detection performance when the pooling operation is replaced with non-maximum suppression (NMS).
More in this topic

Fault Detection

Fault detection identifies abnormal behavior in machines, sensors, infrastructure, software, and autonomous systems before it causes harm or disruption.

See all 104 papers

Explore similar work

CardsList
Mar 13, 2026

Vision-Language Based Expert Reporting for Painting Authentication and Defect Detection

Authenticity and condition assessment are central to conservation decision-making, yet interpretation and reporting of thermographic output remain largely bespoke and expert-dependent, complicating comparison across collections and limiting systematic integration into conservation documentation. Pulsed Active Infrared Thermography (AIRT) is sensitive to subsurface features such as material heterogeneity, voids, and past interventions; however, its broader adoption is constrained by artifact misinterpretation, inter-laboratory variability, and the absence of standardized, explainable reporting frameworks. Although multi-modal thermographic processing techniques are established, their integration with structured natural-language interpretation has not been explored in cultural heritage. A fully automated thermography-vision-language model (VLM) framework is presented. It combines multi-modal AIRT analysis with modality-aware textual reporting, without human intervention during inference. Thermal sequences are processed using Principal Component Thermography (PCT), Thermographic Signal Reconstruction (TSR), and Pulsed Phase Thermography (PPT), and the resulting anomaly masks are fused into a consensus segmentation that emphasizes regions supported by multiple thermal indicators while mitigating boundary artifacts. The fused evidence is provided to a VLM, which generates structured reports describing the location of the anomaly, thermal behavior, and plausible physical interpretations while explicitly acknowledging the uncertainty and diagnostic limitations. Evaluation on two marquetries demonstrates consistent anomaly detection and stable structured interpretations, indicating reproducibility and generalizability across samples.

Eman Ouda, Mohammed Salah, Arsenii O. Chulkov +4

Khalifa University of Science and Technology, Department of Aerospace Engineering, Abu Dhabi, United Arab Emirates · National Research Tomsk Polytechnic University, Russia · Academy of Fine Arts of Naples, Naples, Italy +3

Thermal ImagingAnomaly DetectionMultimodal Anomaly Detection
AskPDF
Aug 29, 2026

Computational Depth Measurement in Thermographic Video: Overcoming Spatial Overfitting via Spatio-Temporal Decoupling

Accurate through-thickness measurement of subsurface delamination depth in Carbon Fiber Reinforced Polymer (CFRP) is important for structural assessment because defect location determines affected load-bearing layers. Optical pulsed thermography (OPT) provides a two-dimensional thermal video rather than volumetric measurements, so depth must be inferred from temporal heat-diffusion responses. A challenge is spatial dataset bias: when calibration defects follow regular grids, regression models may memorize their geometry instead of learning physical relationship between thermal decay and depth. This work introduces a spatio-temporal decoupling architecture that separates spatial defect localization from temporal depth measurement. Defect regions are first localized using segmentation methods, after which thermal responses are spatially averaged and converted into sixteen physics-informed temporal, energy, statistical, and geometric features. These features expose the one-dimensional heat-conduction relationship while withholding pixel coordinates from the depth model. Four regression models are evaluated using specimen-level cross-validation: Random Forest (RF), Gradient Boosting Machine (GBM), Advanced Multi-Layer Perceptron (Adv-MLP), and XGBoost. Unregularized trees and over-parameterized Adv-MLP exhibit calibration collapse under geometric shifts, with errors exceeding 0.5 mm. In contrast, regularized XGBoost with L1/L2 penalties and column sampling maintains cross-specimen calibration, achieving a mean absolute error (MAE) of 0.056 mm and root mean square error (RMSE) of 0.085 mm. Predicted depths are merged with masks to generate Delaunay-triangulated three-dimensional defect models in three to five seconds per specimen. Results show that mathematical regularization and spatio-temporal decoupling reduce spatial memorization in thermal-video depth regression.

Zain Ul Abidin, Habeeban Memon, Junaid Ahmed

Department of Computer Systems Engineering, Sukkur IBA University, Sukkur, Pakistan

Non-Destructive TestingDepth EstimationThermal ImagingIndustrial Inspection
AskPDF
Apr 9, 2026

Data-Driven Automated Identification of Optimal Feature-Representative Images in Infrared Thermography Using Statistical and Morphological Metrics

Infrared thermography (IRT) is a widely used non-destructive testing technique for detecting structural features such as subsurface defects. However, most IRT post-processing methods generate image sequences in which defect visibility varies strongly across time, frequency, or coefficient/index domains, making the identification of defect-representative images a critical challenge. Conventional evaluation metrics, such as the signal-to-noise ratio (SNR) or the Tanimoto criterion, often require prior knowledge of defect locations or defect-free reference regions, limiting their suitability for automated and unsupervised analysis. In this work, a data-driven methodology is proposed to identify images within IRT datasets that are most likely to contain and represent structural features, particularly anomalies and defects, without requiring prior spatial information. The approach is based on three complementary metrics: the Homogeneity Index of Mixture (HI), which quantifies statistical heterogeneity via deviations of local intensity distributions from a global reference distribution; a Representative Elementary Area (REA), derived from a Minkowski-functional adaptation of the Representative Elementary Volume concept to two-dimensional images; and a geometrical-topological Total Variation Energy (TVE) index, also based on two-dimensional Minkowski functionals, designed to improve sensitivity to localized anomalies. The framework is validated experimentally using pulse-heated IRT data from a carbon fiber-reinforced polymer (CFRP) plate containing six artificial defects at depths between 0.135 mm and 0.810 mm, and is further supported by one-dimensional N-layer thermal model simulations. The results demonstrate robust and unbiased ranking of image sequences and provide a reliable basis for automated defect-oriented image selection in IRT.

Harutyun Yagdjian, Martin Gurka

Leibniz-Institut für Verbundwerkstoffe GmbH, Erwin-Schrödinger-Straße 58, 67663 Kaiserslautern

Non-Destructive TestingThermal ImagingData SelectionIndustrial Inspection
AskPDF