In plain words: A neural network turns shadowgraph images—shadows cast when light bends through changing density—straight into full density maps, guided by physics rules so answers stay sensible. It does the job in one pass, instead of the usual separate image-processing and equation-solving steps.
Abstract · Physics-informed Shadowgraph Network: An End-to-end Density Field Reconstruction Method
This study presents a novel approach for quantificationally reconstructing density fields from shadowgraph images using physics-informed neural networks
Xutun Wang, Yuchen Zhang, Zidong Li, Haocheng Wen, Bing Wang
arXiv:2410.20203 · physics.flu-dyn, cs.AI · submitted Oct 26, 2024 · updated Nov 2, 2024
abstract · pdf · Experimental Thermal and Fluid Science (2025)
