Principal component analysis guided autoencoding for structured dimensionality reduction in active infrared thermography

Active Infrared thermography (AIRT) is a widely adopted non-destructive testing (NDT) technique for detecting subsurface anomalies in industrial components. Due to the high dimensionality of AIRT data, current approaches employ autoencoders (AEs) for dimensionality reduction. However, the latent space learned by AIRT AEs lacks consistency and structure, limiting their effectiveness in downstream defect analysis tasks. To address this limitation, this paper proposes principal component analysis (PCA)-guided autoencoding framework for structured dimensionality reduction in AIRT. The proposed PCA-Guided AE captures intricate, non-linear features in thermographic signals while enforcing a structured latent space. A novel loss function, PCA distillation loss, is introduced to guide AEs to align the latent representation with structured PCA components, while capturing the intricate, non-linear patterns in thermographic signals. To evaluate the utility of the learned, structured latent space, we propose a neural network–based evaluation metric that assesses its suitability for artificial intelligence (AI) based defect analysis. Experimental results show that the proposed PCA-guided AE outperforms state-of-the-art dimensionality reduction methods on carbon-fiber reinforced polymers (CFRP), polylactic acid (PLA), and polyvinyl chloride (PVC) samples in terms of contrast, signal-to-noise ratio (SNR), and neural network-based metrics. Code