Speaker
Description
Laser speckle, a phenomenon resulting from the scattering of coherent light from a rough surface, is rich with information encoding the surface's micro-topography. This work presents a novel optical methodology that decodes this information for quantitative, non-contact surface texture analysis from a single, static speckle. Our approach bypasses the temporal and mechanical constraints of traditional scanning profilometry by establishing a direct, quantitative correlation between the statistical properties of the speckle intensity field and surface texture parameters. The technique was experimentally confirmed using a set of industrially significant composite metal-ceramic samples. An optical configuration, comprising a He-Ne laser and a CMOS sensor, was used to capture far-field speckle patterns. Through rigorous cross-validation analysis, we demonstrate a strong correlation between speckle statistics and functional surface descriptors. Specifically, our model shows strong predictive power for Fractal Dimension and Gray-Level Co-occurrence Matrix (GLCM) Contrast, and accurately predicts surface Kurtosis. The sub-second acquisition-to-analysis pipeline underscores the technique's potential for high-throughput applications. This research validates a powerful, single-image optical approach, paving the way for compact, mechanically robust optical instruments for real-time, in-situ surface metrology.
| Type of presence | Presence at Taras Shevchenko National University |
|---|
