LAUSR

As a new imaging inspection method with characteristics of a wide view field and non-contact, spatial frequency domain imaging (SFDI) is very suitable to evaluate the optical properties of agricultural products to ensure the sustainable development of agriculture. However, due to the unique forward scattering characteristics of fruit skin, only a few photons can return to the skin surface after interacting with the flesh, thus affecting the detection accuracy of the flesh layer. This study aims to propose a more accurate and wider applicable method to extract the optical properties of two-layer tissue from SFDI measurements. Firstly, a two-layer model was proposed by optimizing the reflectivity of the flesh layer through the optical properties and thickness of the skin layer. Secondly, the influence of the optical properties and thickness of different skin layers on the reflectivity optimization of the flesh layer was investigated by a Monte Carlo simulation, and then, the accuracy and effectiveness of the proposed model was evaluated for practical inspection by phantom experiments. Finally, this model was used to obtain the optical properties, layer by layer, of four thin-skinned fruits (pear, apple, peach and muskmelon) to verify its universality. The results showed that, for the skin layer, the average errors of the absorption coefficient (μa1) and the reduced scattering coefficient (μ′s1) were 10.87% and 7.91%, respectively, and for the flesh layer, the average errors of the absorption coefficient (μa2) and the reduced scattering coefficient (μ′s2) were 16.76% and 8.64%, respectively. This study provides the basis for the SFDI detection of optical properties of two-layer tissue such as thin-skinned fruits, which can be further used for nondestructive fruit quality evaluations.