Sun Ye, Huang Yuping, Pan Leiqing, Wang Xiaochan
College of Engineering, Nanjing Agricultural University, Nanjing 210031, China.
College of Mechanical and Electronic Engineering, Nanjing Forestry University, Nanjing 210037, China.
Foods. 2021 Feb 10;10(2):388. doi: 10.3390/foods10020388.
The main objective was to measure the optical coefficients of peaches after bruising at different maturity levels and detect bruises. A spatially resolved method was used to acquire absorption coefficient (μ) and the reduced scattering coefficient (µ') spectra from 550 to 1000 nm, and a total of 12 groups (3 maturity levels * 4 detection times) were used to assess changes in µ and µ' resulting from bruising. Maturation and bruising both caused a decrease in µ' and an increase in µ, and the optical properties of immature peaches changed more after bruising than the optical properties of ripe peaches. Four hours after bruising, the optical properties of most samples were significantly different from those of intact peaches ( < 0.05), and the optical properties showed damage to tissue earlier than the appearance symptoms observed with the naked eye. The classification results of the Support Vector Machine model for bruised peaches showed that μ had the best classification accuracy compared to μ' and their combinations (µ × µ', µ). Overall, based on μ, the average detection accuracies for peaches after bruising of 0 h, 4 h, and 24 h were increased.
主要目的是测量不同成熟度的桃子在 bruised 后的光学系数并检测 bruised 情况。采用空间分辨方法获取 550 至 1000 nm 的吸收系数（μ）和约化散射系数（µ'）光谱，共使用 12 组（3 个成熟度水平×4 个检测时间）来评估 bruised 导致的 μ 和 µ' 的变化。成熟和 bruised 均导致 µ' 降低和 μ 增加，未成熟桃子 bruised 后的光学特性变化比成熟桃子的光学特性变化更大。bruised 后 4 小时，大多数样品的光学特性与完整桃子的光学特性有显著差异（<0.05），并且光学特性比肉眼观察到的外观症状更早显示出组织损伤。bruised 桃子的支持向量机模型分类结果表明，与 µ' 及其组合（µ×µ'，µ）相比，μ 具有最佳分类准确率。总体而言，基于 μ，bruised 后 0 小时、4 小时和 24 小时桃子的平均检测准确率有所提高。
