We proposed and demonstrated a digital method of dispersion compensation suitable for spectral-domain optical coherence tomography. The wavelength coordinate of the coherence spectrum was calibrated digitally using a two-order polynomial. A software-based scheme was introduced to determine the polynomial coefficients of the polynomial fitting spectrum wavelength. Therefore, the spectrum deformation introduced by dispersion can be compensated effectively. This method was experimentally validated by in vivo imaging an early-stage chick embryonic heart.
In this paper an attempt ofNIR analysis to measure the milk constituents is introduced. In milk, light is absorbed and scattered simultaneously. The concentrations of each component no longer have linear relationship to the absorbency, but proportional to the absorption coefficient and scattering coefficient. It is necessary to separate the effect of absorption and scattering. Compare with the widely used NIR method, we do not calibrate the prediction model by diffuse reflectance. The calibration is composed by two steps. Firstly, the optical parameters are measured by double integrating sphere; secondly, calibration is made by optical parameters. Concentration of protein and fat are within reasonable range. Fifty samples are acquired for calibration. The experimental result shows a higher prediction accuracy of protein than the intensity calibration method, and a comparable accuracy of fat.