【摘要】 为探究软枣猕猴桃采后后熟过程中可溶性固形物含量（soluble solids content, SSC）变化和分布规律，利用高光谱成像结合化学计量学方法实现其SSC无损检测与可视化。首先，采集25℃下不同贮藏天数软枣猕猴桃的高光谱数据，并测定其SSC。其次，采用不同预处理方法对光谱数据进行处理，确定最佳预处理方法；然后，基于3种特征波段提取方法优选特征波段，构建偏最小二乘回归（partial least squares regression, PLSR）、极限学习机（extreme learning machine, ELM）和粒子群优化的极限学习机（particle swarm optimization-extreme learning machine, PSO-ELM）可溶性固形物含量预测模型。结果表明：基于竞争性自适应重加权采样算法（competitive adaptive reweighted sampling, CARS）提取特征波长的PSO-ELM模型的预测效果最佳，测试集Rp2为0.934,RMSEP为0.952,RPD为2.277。最后，基于CARS-PSO-ELM模型计算软枣猕猴桃每个像素点的SSC，生成可视化分布图，直观地呈现出不同贮藏天数软枣猕猴桃SSC变化的空间分布特征，为软枣猕猴桃的品质评价和贮运销售提供重要参考。更多还原

【Abstract】 In order to explore the change and distribution of soluble solids content(SSC) of Actinidia argute during the post-ripening,hyperspectral imaging technology combined with chemometrics was utilized to achieve non-destructive detection and visualization of SSC in this study. Firstly, hyperspectral imager(400-1 000 nm) was used to obtain the hyperspectral data of Actinidia argute stored at25 ℃ for different storage days, and their SSC value were measured. Secondly, the spectral data of Actinidia argute was preprocessed by different preprocessing methods, and the optimal pretreatment method was determined. Thirdly, the characteristic wavelengths were extracted by three methods, and they were taken as input, the SSC prediction models based on partial least squares regression(PLSR),extreme learning machine(ELM) and particle swarm optimization-extreme learning machine(PSO-ELM) were established respectively. The results showed that the PSO-ELM model which extracted characteristic wavelength through competitive adaptive reweighted sampling(CARS) method performed better than the other models. The CARS-PSO-ELM model achieved the optimal prediction accuracy with the Rp2of 0.934, the RMSEP of 0.952, and the RPD of 2.277. Finally, based on the optimal model, the SSC of each pixel of Actinidia argute was calculated, and the visual distribution map was generated, which intuitively showed the spatial distribution characteristics of the change of SSC in Actinidia argute at different storage days, and provided an important reference for the quality evaluation, storage, transportation and sales of Actinidia argute.更多还原