Influence of the size and skin thickness of apple varieties on the retrieval of internal optical properties using Vis/NIR spectroscopy: A Monte Carlo-based study

Visible/near-infrared spectroscopy is a well-established method to measure optical properties of tissues, assuming that a light propagation model can be used to recover absorption and reduced scattering coefficients from non-invasive probing. Spectroscopic measurements have achieved success in non-destructive assessment of apple optical properties and quality attributes. However, the spectroscopy of apples must consider the size of the fruit and the presence of the thin skin layer that surrounds the flesh, to correctly read the signals acquired on the boundary. In this research, the fruit was modelled as a two layer spherical structure with various radii and finite thickness of the upper skin layer. Monte Carlo computations were performed to generate time-resolved reflectance and spatially-resolved reflectance measurements. Simulated data were then fitted using a procedure based on Levenberg Marquardt algorithm with specific semi-infinite models. The errors in the retrieved optical properties of the flesh (absorption coefficient mu(a), and reduced scattering coefficient mu'(s)) were studied as functions of apple radius, skin thickness, and source detector distance, for given optical parameter sets assigned to the flesh and the skin. The results suggest that the time-resolved reflectance spectroscopy may probe optical properties of the flesh regardless of the skin layer, when a sufficient source detector distance (15 mm) is used for the measurements. Similar results were found in case of using the spatially resolved spectroscopy, because measurements extend up to 15-29 mm by steps of 1 mm or 2 mm. The computations also show that the curvature of the boundary has noticeable effect on the errors in the retrieved optical coefficients of the flesh. However, results from time-resolved spectroscopy are more influenced by the size of apples, compared with the spatially-resolved spectroscopy. (C) 2015 Elsevier B.V. All rights reserved.