Spatial frequency domain imaging coupled with frequency optimization for estimating optical properties of two-layer food products

Authors: HU, DONG - Zhejiang A & F University; Lu, Renfu; YING, YIBIN - Zhejiang University

Submitted to: ASABE Annual International Meeting
Publication Type: Proceedings
Publication Acceptance Date: 7/22/2019
Publication Date: 7/22/2019
Citation: Hu, D., Lu, R., Ying, Y. 2019. Spatial frequency domain imaging coupled with frequency optimization for estimating optical properties of two-layer food products. In: Proceedings of the ASABE Annual International Meeting, July 7-10, 2019, Boston, Massachusetts. Paper No. 1900695.

Interpretive Summary: Measurement of optical absorption and scattering properties can provide important information about quality and composition of horticultural and food products. Many food products such as fruit, are composed of a thin layer of skin encompassing the flesh, whose physical and chemical properties are distinctly different. Hence, it would be advantageous to measure the optical properties of each layer in order to better evaluate quality and composition of food products. In this research, spatial frequency domain reflectance technique, an emerging optical technique, was used for optical property measurement of food products that are composed of two homogeneous layers. In the conventional approach, the optical properties of layered materials are estimated from the measured reflectance using the so-called single step method, in which all parameters are estimated at once. This approach, however, is not satisfactory because of large, unacceptable errors for estimating each layer. In this research, we proposed a step-wise method, coupled with the optimized frequency range, in which the optical properties of each layer are estimated in separate steps, so as to reduce the overall estimation errors in the optical property measurement of layered food products. Through computer simulations and experimental validations with real samples, we demonstrated that the proposed method, coupled with the optimized frequency range, achieved significant improvements in accuracy (in the range of 53% to 63%) over the conventional one-step method for the measurement of absorption coefficients of the first and second layers as well as the scattering coefficient of second layer. The proposed method was used for estimating optical properties for the skin and flesh of apple fruit for four cultivars, and it compared favorably with a reference method. The proposed stepwise method with the optimized frequency range enables better measurement of the optical properties of layered food products for spatial frequency domain reflectance technique.

Technical Abstract: This research was aimed at optimizing the frequency region through an inverse algorithm for better quantification of the optical absorption and reduced scattering coefficients of two-layer food products from spatial frequency domain reflectance. The frequency region, defined by start and end frequencies, was first optimized for parameter estimations of the first and second layers, respectively. Estimation accuracies were then validated by comparing with the conventional all-at-once method through Monte Carlo simulations. By using the optimized frequency range, the average accuracies for estimation of the absorption coefficients of first and second layers and the reduced scattering coefficient of second layer were improved by 52.9%, 63.0% and 62.1%, respectively. No improvement for the estimated the scattering coefficient of first layer was found because its mean absolute error was already very low (i.e., 2.4%) and well within the acceptable level. Experimental results for two-layer solid phantoms and liquid milk samples in the wavelengths of 650-830 nm further validated the effectiveness of the stepwise method with the optimized frequency region. Finally, the stepwise method, coupled with the optimized frequency region, was used to estimate the optical properties of skin and flesh of apples for four cultivars (i.e., Delicious, Golden Delicious, Jonagold and Red Rome). The results were compared with those obtained using the single integrating sphere technique.