The surface roughness prediction model based on a support vector machine was proposed and the multi-wavelength fiber optic sensor was established. The specimens with different surface roughness selected as the test samples were analyzed by using the prediction model when the incident wavelengths were 650 nm and 1310 nm, respectively. The working distance of 2.5 mm ~3.5 mm was chosen as the optimum measurement distance. The experimental results indicate that the error range of surface roughness is 0.74% ~7.56% at 650 nm, and the error range of surface roughness is 1.03% ~5.92% at 1310 nm. The average relative error is about 2.669% at 650 nm, while it is about 2.431% at 1310 nm. The error of roughness measurement is less than 3% by using the model, which is acceptable. The error of surface roughness based on the prediction model is smaller than that by using the characteristic curves between surface roughness and the scattering intensity ratio.
© 2016 Optical Society of America
OSA Recommended Articles
Equations on this page are rendered with MathJax. Learn more.