Unlocking freeform structured surface denoising with small sample learning: Enhancing performance via physics-informed loss and detail-driven data augmentation

• Published in: Advanced engineering informatics Vol. 62; p. 102733
• Main Authors: Cui, Weixin, Lou, Shan, Zeng, Wenhan, Kadirkamanathan, Visakan, Qin, Yuchu, Scott, Paul J., Jiang, Xiangqian
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.10.2024
• Subjects: Denoising; Detail-driven data augmentation; Physics-informed loss; Small sample training; Surface metrology; Small sample training; Detail-driven data augmentation; Physics-informed loss; Surface metrology; Denoising
• ISSN: 1474-0346; 1873-5320
• DOI: 10.1016/j.aei.2024.102733

Denoising plays a vital role in freeform structured surface metrology. Traditional techniques, such as Gaussian and partial differential equation-based diffusion filters, often involve a time-consuming calibration process, particularly for complex surfaces. The main challenge lies in automating the denoising operation while accurately preserving features for varied surface textures. To address this challenge, an automatic approach PI-DnCNN based on small sample learning is presented in this paper. Denoising convolutional neural network (DnCNN) is employed as the basic architecture of this approach, due to its effectiveness in tackling mix-level Gaussian noise and adapting to small training datasets. Acknowledging the constraints of limited datasets, a novel physics-informed denoising loss function marrying filtering techniques is proposed to improve model performance. Additionally, a hybrid data augmentation strategy is developed to enhance the recognition of complex components. The paper also reports a set of experiments to demonstrate the presented approach in terms of performance over conventional techniques, enhancements with limited sample sizes, and applicability in general image denoising. The experiment results suggest that the presented approach consistently achieves higher average scores compared to traditional filters and emerges superior compared to the conventional DnCNN loss across different dataset sizes. In addition, the proposed loss also shows effectiveness in general image denoising, which suggests the robustness and universality of the approach. •Develop an automatic surface denoising method that surpasses conventional filters.•Present a physics-informed loss function that improves denoising with limited data.•Propose a hybrid data augmentation strategy that enhances model generalisation.