The growth characteristics of Aspergillus parasiticus incubated on two culture media were examined using shortwave infrared (SWIR, 1000–2500 nm) hyperspectral imaging (HSI) in this work. HSI images of the A. parasiticus colonies growing on rose bengal medium (RBM) and maize agar medium (MAM) were recorded daily for 6 days. The growth phases of A. parasiticus were indicated through the pixel number and average spectra of colonies. On score plot of the first principal component (PC1 T and PC2, four growth zones with varying mycelium densities were identified. Eight characteristic wavelengths (1095, 1145, 1195, 1279, 1442, 1655, 1834 and 1929 nm) were selected from PC1 loading, average spectra of each colony as well as each growth zone. Furthermore, support vector machine (SVM) classifier based on the eight wavelengths was built, and the classification accuracies for the four zones (from outer to inner zones) on the colonies on RBM were 99.77%, 99.35%, 99.75% and 99.60% and 99.77%, 99.39%, 99.31% and 98.22% for colonies on MAM. In addition, a new score plot of PC2 and PC3 was used to differentiate the colonies incubated on RBM and MAM for 6 days. Then characteristic wavelengths of 1067, 1195, 1279, 1369, 1459, 1694, 1834 and 1929 nm were selected from the loading of PC2 and PC3. Based on them, a new SVM model was developed to differentiate colonies on RBM and MAM with accuracy of 100.00% and 99.99%, respectively. In conclusion, SWIR hyperspectral image is a powerful tool for evaluation of growth characteristics of A. parasiticus incubated in different culture media.

In this paper, a methodology based on characteristic spectral bands of near infrared spectroscopy (1000–2500 nm) and multivariate analysis was proposed to identify camellia oil adulteration with vegetable oils. Sunflower, peanut and corn oils were selected to conduct the test. Pure camellia oil and that adulterated with varying concentrations (1–10% with the gradient of 1%, 10–40% with the gradient of 5%, 40–100% with the gradient of 10%) of each type of the three vegetable oils were prepared, respectively. For each type of adulterated oil, full-spectrum partial least squares partial least squares (PLS) models and synergy interval partial least squares (SI-PLS) models were developed. Parameters of these models were optimized simultaneously by cross-validation. The SI-PLS models were proved to be better than the full-spectrum PLS models. In SI-PLS models, the correlation coe±cients of predition set (Rp) were 0.9992, 0.9998 and 0.9999 for adulteration with sunflower oil, peanut oil and corn oil seperately; the corresponding root mean square errors of prediction set (RMSEP) were 1.23, 0.66 and 0.37. Furthermore, a new generic PLS model was built based on the characteristic spectral regions selected from the intervals of the three SI-PLS models to identify the oil adulterants, regardless of the adultrated oil types. The model achieved with Rp 0.9988 and RMSEP=1.52. These results indicated that the characteristic near infrared spectral regions could determine the level of adulteration in the camellia oil.

山茶油素有“东方橄榄油”美誉, 实现掺假山茶油的鉴别具有重要实用价值, 采用近红外光谱技术对掺有葵花油的山茶油进行检测。 分别以1%, 5%, 10%为梯度制备掺假比例不同的山茶油样品, 并根据掺假比例将其分为A组（0%～5%）和B组（6%～10%）共11个样品, C组（15%～40%）6个和D组（50%～100%）6个样品。 将每个掺假样品充分混匀后再分为9份, 依次采集其1 000～2 500 nm范围的吸收光谱, 共获得207条光谱曲线。 每组样品的光谱数据按2∶1随机分为训练集与验证集。 经去除首尾噪声后, 通过主成分分析法（principal component analysis, PCA）降维, 并利用前四个主成分建立了鉴别山茶油不同掺假等级的主成分-支持向量机判别模型, 训练集与验证集的总体判别准确率分别达96.38%和94.20%; 进一步, 通过对前四个主成分的载荷系数的分析, 并结合原始光谱, 提取建模过程中权重较大的波长并解析其化学含义, 最终确定出五个特征波长: 1 212, 1 705, 1 826, 1 905及2 148 nm, 以此波长重新建立近红外特征光谱山茶油掺假等级判别模型, 对训练集与验证集的总体判别准确率也达到了94.20%和92.75%。 研究结果表明, 利用近红外光谱和特征光谱均能够较好实现山茶油掺假等级的鉴别, 同时所建立的近红外特征光谱模型也为设计相应的掺假山茶油实用便携式检测仪器提供了理论基础。

利用波长范围在833～2 500 nm的傅里叶变换近红外光谱(Fourier transform near infrared spectroscopy, FT-NIR)对不同霉变程度的玉米颗粒进行检测区分。 首先, 为避免光谱数据首尾噪声影响, 对比四种常见的预处理方法, 最终选择移动平均平滑法对原始光谱数据进行预处理; 然后为选出合适的样本集划分方法以提高模型预测性能, 对常见的四种方法进行对比, 最终利用SPXY(sample set partitioning based on joint x-y distance)法进行样本集划分; 进一步为减少数据量, 降低维度, 使用连续投影算法(successive projections algorithm, SPA)提取出7个特征波长, 分别为833, 927, 1 208, 1 337, 1 454, 1 861和2 280 nm; 最后, 将七个特征波长数据作为输入, 选取径向基函数(radial basis function, RBF)作为支持向量机(support vector machine, SVM)核函数, 取参数C=7 760 469, γ=0.017 003建立判别模型。 SVM模型对训练集和测试集的预测准确率分别达到97.78%和93.33%。 另取不同品种的玉米颗粒, 以同样的标准挑选样品组成独立验证集, 所建立的判别模型对独立验证集的预测准确率达到91.11%。 结果表明基于SPA和SVM能有效地对玉米颗粒霉变程度进行判别, 所选取的7个特征波长为实现在线霉变玉米颗粒近红外检测提供了理论依据。

黄曲霉毒素是广泛存在于玉米中且具有剧毒的一种代谢产物, 以美国农业部农业研究署(USDA-ARS) Toxicology and Mycotoxin Research Unit提供的2010年先锋玉米为研究对象, 验证了高光谱成像技术对玉米中黄曲霉毒素检测的可行性。 以甲醇为溶剂制备四种不同浓度的黄曲霉毒素溶液, 并将其逐一滴在等量的4组共120粒玉米颗粒表面, 以未处理的30粒洁净玉米作为一组对照样本, 将大小、 形状相似的150个样品随机分为训练集103个, 验证集47个; 对获取的400～1 000 nm波段范围内的高光谱图像, 先进行标准正态变量变换(standard normal variate transformation, SNV)预处理, 然后引入基于Fisher判别最小误判率的方法选择最优波长, 并以所选波长作为Fisher判别分析法的输入建立判别模型, 对玉米颗粒表面不同浓度的黄曲霉毒素进行识别, 最后对模型判别正确率进行了验证。 结果表明, 选取四个最优波长(812.42, 873.00, 900.36和965.00 nm)时Fisher判别分析模型对训练集与验证集的准确率分别为87.4%和80.9%。 该方法为含黄曲霉毒素玉米颗粒便携式检测仪器的开发, 以及对田间霉变玉米自然代谢产生毒素的检测奠定了技术基础。