Metal thin film photoacoustic signal zero point measurement method, device and equipment and storage medium
By combining a physical model of ultrasound generation and propagation with a trained neural network, the accuracy and stability issues of zero-point measurement of photoacoustic signals in thin metal films were solved, achieving high-precision, automated zero-point positioning and film thickness calculation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SKYVERSE TECH CO LTD
- Filing Date
- 2026-03-27
- Publication Date
- 2026-07-21
AI Technical Summary
Existing technologies suffer from insufficient accuracy and stability when determining the time zero point of photoacoustic signals from thin-film metals. This is especially true when the signal-to-noise ratio is low or the features are not obvious. Manual or semi-automatic identification methods have poor repeatability, while data-driven methods lack physical constraints and are not adaptable enough.
A physical model of ultrasound generation and propagation under femtosecond laser action was established, the zero-point search time window was calculated, and a pre-trained zero-point recognition neural network was used to accurately locate the zero point using photoacoustic reflectivity time series signals. The method combines physical model constraints and neural network recognition.
High-precision and stable zero-point positioning was achieved, which improved the adaptability of the method under different measurement conditions and the physical interpretability of the results, providing a reliable foundation for subsequent film thickness calculation.
Smart Images

Figure CN122430249A_ABST