| Knowledge of optical properties is critical for quantitative understanding of light interaction with food materials and the development of effective optical techniques for quality evaluation and safety inspection of food and agricultural products. Currently, there are no appropriate methods available for rapid and reliable measurement of the optical properties of horticultural and food materials. Hence the objective of this research was to develop an optical technique for fast and noncontact measurement of the optical properties of turbid food materials, quantify light propagation in apples, and relate optical properties to the chemical constituents and quality attributes of food products.A hyperspectral imaging system in line scanning mode was designed, calibrated and evaluated for acquiring spatially resolved diffuse reflectance images from turbid food materials over the visible and short-wave near-infrared region (500-1000 nm). Algorithms were developed based on a steady-state diffusion theory model for determining absorption and reduced scattering coefficients (mua and mus', respectively) from the spatially resolved diffuse reflectance profiles. Validation tests on liquid simulation samples showed that the hyperspectral imaging method provided good measurement of mua and mus' with average errors of 16% and 11%, respectively.The optical properties (i.e., mua and mus') of liquid foods (citrus juice, grapefruit juice, MinutemaidRTM) pulp free orange juice, TropicanaRTM pulp free orange juice, mixed orange-pineapple juice, V8RTM vegetable juice, and milk) and horticultural crops (apple, peach, pear, kiwifruit, plum, cucumber, zucchini squash, and tomato) were determined over the spectral regions of 530-900 nm and 500-1000 nm, respectively. Values of mua and mus' for the test juice samples, milk samples, and fruit and vegetable samples were in the range of 0.0-0.8 cm -1 and 3-20 cm-1, 0.0-2.6 cm-1 and 15-50 cm-1, and 0.0-0.5 cm-1 and 8-25 cm-1 , respectively. Absorption spectra of horticultural crops were featured by major pigments (i.e., chlorophyll, anthocyanin, lycopene, and carotenoid) and water in the sample tissues, and scattering spectra steadily decreased with the increase of wavelength.Monte Carlo simulation models were developed to investigate light propagation in apple fruit using typical mua and mus' values of 'Golden Delicious' apples. Simulation results suggested that the optimum spatial sensing range under the present imaging system setup was about 13 mm for apples. The 1% light penetration depths estimated using the optical properties ranged from 0.91 to 5.78 cm for the fruit and vegetable samples over the wavelength range of 500-1000 nm. Pigments and water in the plant tissues determined the major spectral features of penetration depths.Values of mua and mus' at 600 nm for the milk samples were highly correlated to the fat content with the correlation coefficient (r) of 0.995 and 0.998, respectively. Absorption patterns for tomatoes changed with the stage of their ripeness, and values of mus' decreased with the softening of tomatoes. The combined data of mua and mus' for 'Golden Delicious' apples were able to predict fruit firmness with r = 0.88 and the standard error of prediction (SEP) of 5.66 N, and soluble solids content with r = 0.82 and SEP = 0.75%.The hyperspectral imaging technique is fast, noncontact and easy in use, making it especially suitable for measurement of the optical properties of turbid food products. |
