Method and system for identifying defects in a film based on machine vision
By acquiring polarization images of thin films using multi-wavelength pulsed polarized light, constructing interlayer interface energy dissipation fingerprints, inverting viscoelastic constitutive parameters, and predicting fatigue life, this method solves the problem of not being able to identify latent interlayer interface weakening in existing technologies, and achieves efficient detection and process optimization of thin films.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANGHAI ASTRACE NEW MATERIAL TECH CO LTD
- Filing Date
- 2026-05-15
- Publication Date
- 2026-07-17
AI Technical Summary
Existing machine vision inspection technology cannot perceive the dynamic energy dissipation process of interlayer interfaces, cannot invert the viscoelastic constitutive parameters of interlayers, cannot predict the fatigue crack initiation lifetime, and cannot identify the weakened regions of latent interlayer interfaces, leading to functional failure of the thin film during use.
By projecting multi-wavelength pulsed polarized light onto the thin film, acquiring time-series polarization images, extracting the polarization state evolution trajectory of pixels, constructing an interlayer interface energy dissipation fingerprint spectrum, inverting the storage modulus and loss factor distribution map, combining with a fatigue mechanics model to calculate fatigue life, and identifying latent interlayer interface weakening regions.
It enables online sensing of interfaces that have not yet cracked but whose viscoelasticity has degraded, provides inversion of the elastic modulus and loss factor of interlayer interfaces, predicts fatigue crack initiation lifetime, locates the weakening zone of latent interlayer interfaces, and supports long-term durability evaluation and process optimization of thin films.
Smart Images
Figure CN122409658A_ABST