Cover story: Yin W, Che YX, Li XS et al. Physics-informed deep learning for fringe pattern analysis. Opto-Electron Adv 7, 230034 (2024).

Deep learning has yielded transformative success across optics and photonics, especially in optical metrology. Deep neural networks (DNNs) with a fully convolutional architecture (e.g., U-Net and its derivatives) have been widely implemented in an end-to-end manner to accomplish various optical metrology tasks, such as fringe denoising, phase unwrapping, and fringe analysis. However, the task of training a DNN to accurately identify an image-to-image transform from massive input and output data pairs seems at best naïve, as the physical laws governing the image formation or other domain expertise pertaining to the measurement have not yet been fully exploited in current deep learning practice. Recently, the research group of Prof. Chao Zuo at the Smart Computational Imaging Laboratory (SCILab) of Nanjing University of Science and Technology (NJUST) reported a physics-informed deep learning technique for fringe pattern analysis (PI-FPA). By embedding the prior knowledge of traditional single-frame fringe pattern analysis methods in the DNN, PI-FPA utilizes the learnable adaptive filtering in the Fourier transform domain to directly recover reliable initial phases from the single-frame fringe, and guides the DNN to achieve high-accuracy and computationally efficient phase measurement. This research presents that the synergy of physics-priors-based traditional tools and data-driven learning approaches opens new avenues for fast and accurate single-shot 3D shape measurement of dynamic scenes.