A Dual-Stage Framework For Cavity Detection In Nuclear Materials Using Image Super-Resolution (ISR) And Yolov10s

Abstract

Radiation-induced cavities and voids within structural materials present a significant challenge for ensuring the long-term safety and efficiency of nuclear reactors. Accurate detection and quantification of these cavities are critical for understanding irradiation damage mechanisms and predicting material degradation. Traditional object detection models such as YOLOv8 and Faster R-CNN have shown reasonable performance in cavity detection; however, their bounding-box-based outputs often fail to precisely localize small or irregularly shaped cavities, particularly under degraded imaging conditions. To address these limitations, this paper introduces CavityNet, a novel two-stage deep learning framework that combines Enhanced Super-Resolution Generative Adversarial Network (ESRGAN) with a Swin Transformer-based UNet (Swin-UNet) architecture for improved cavity detection in irradiated microstructures. In the first stage, ESRGAN enhances underfocused or overfocused scanning transmission electron microscopy (S/TEM) images by reconstructing fine textures and high-frequency cavity boundaries. In the second stage, these super-resolved images are passed through a Swin-UNet model, which integrates hierarchical self-attention mechanisms with a UNet-style encoder-decoder pipeline to achieve pixel-level segmentation of cavities. This approach enables both global contextual understanding and local feature refinement, improving detection accuracy in complex microstructural environments. The model is trained and evaluated using publicly available datasets from the Canadian Nuclear Laboratory (CNL) and Nuclear Oriented Materials & Examination (NOME), achieving an F1-score improvement of 31% on underfocused images and a precision improvement of 161.3% with ISR augmentation. The proposed CavityNet framework represents a significant step forward in automated microstructural analysis, offering enhanced performance, robustness to imaging conditions, and potential for real-time material degradation monitoring in nuclear environments.