Skip to main content
ARS Home » Midwest Area » East Lansing, Michigan » Sugarbeet and Bean Research » Research » Publications at this Location » Publication #240306
Title: Two-layer Model for Measuring the Optical Properties of Turbid Materials Based on Spatially Resolved Hyperspectral Diffuse Reflectance Images

Author
item CEN, HAIYAN - Michigan State University
item Lu, Renfu

Submitted to: Inverse Problems Symposium
Publication Type: Abstract Only
Publication Acceptance Date: 5/20/2009
Publication Date: 5/31/2009
Citation: Cen, H., Lu, R. 2009. Two-layer Model for Measuring the Optical Properties of Turbid Materials Based on Spatially Resolved Hyperspectral Diffuse Reflectance Images [abstract]. Inverse Problems Symposium.

Interpretive Summary:

Technical Abstract: Hyperspectral imaging-based spatially resolved technique is useful for determining the optical properties of fruits and food products that are homogenous. To better characterize fruit properties and quality attributes, it is desirable that fruit be considered as two homogeneous layers, i.e., skin and flesh. This research was aimed at developing a nondestructive method to determine the spectral absorption and scattering properties of two-layer turbid materials with characteristics of fruit. A two-layer diffusion model was used to describe light propagation in layered biological materials. An inverse algorithm was developed for extraction of optical properties from the spatially resolved diffuse reflectance acquired by a hyperspectral imaging system. Sensitivity coefficients were calculated to analyze the response of reflectance to perturbations in values of optical parameters. Monte Carlo (MC) simulations and experimental data from two model samples with known optical properties were performed to validate the model and inverse algorithm. The sensitivity coefficients of the optical parameters were relatively large and uncorrelated, which were desirable conditions for estimating the parameters. The differences of reflectance calculated by the two-layer diffusion model and MC simulations were less than 6% for source-detector distance greater than 1.5 mm. It indicated that the diffusion model accurately quantified light propagation in two-layer turbid media. The average errors for determining two and four optical parameters from MC generated reflectance data were 6.8% and 15.3%, respectively. The reflectance profiles extracted from the hyperspectral images of the two model samples at selected wavelengths were in good agreement with the theoretical predictions that were calculated from the diffusion model. The errors in estimating the absorption and reduced scattering coefficients for the first or top layer of the two model samples were 11.3-23.0% and 3.8-18.4% respectively, which were larger than those obtained from MC simulations due to greater uncertainties in the experimental data. This research demonstrated the feasibility of evaluating the optical properties of two-layer turbid materials, characteristic of the skin and flesh of fruit, using a two-layer diffusion model for spatially resolved reflectance acquired by a hyperspectral imaging system. Further research is needed to optimize the hyperspectral imaging system and improve the inverse algorithm in order to achieve more accurate reflectance measurement for two-layer turbid materials.