基于透射光谱的苹果霉心病多因子无损检测

针对目前苹果霉心病难以检测的问题，提出一种基于透射光谱的苹果霉心病多因子无损检测方法，通过融合多波段透射光谱与苹果直径，构建苹果霉心病判别模型，实现了苹果霉心病无损检测。搭建光谱测试范围在200～1 025 nm的透射光谱采集平台，实验获取232 个苹果样本的透射光谱数据，采用游标卡尺获得苹果直径数据。采用杂散光校正，非线性校正对苹果透射原始光谱进行预处理，选取与霉心病发病相关的12 个波段透射光强值，结合苹果的直径进行主成分分析，将分析的结果作为自变量，建立苹果霉心病Fisher判别模型。经过异校验验证，模型总体识别率为93.1%，而仅采用透射光谱构建的模型识别率为91.37%。结果表明，基于透射光谱与直径结合的多因子检测方法可实现苹果霉心病的准确判定，为苹果霉心病的快速、无损检测提供可行思路。

Multiple-Factor Nondestructive Detection of Moldy Core in Apples Based on Transmission Spectra

Currently, the detection of moldy core in apples is still a problem. This study aimed to develop a method for
multiple-factor nondestructive detection of moldy core in apples based on transmission spectra. We built a discriminant
model which enables nondestructive detection of moldy core in apples based on both multiple-band transmission spectrum
and apple diameter. A spectrum acquisition platform was constructed to acquire the transmission spectra (200–1 025 nm) of
232 apples. and their diameters were measured with a vernier caliper. Stray light correction and nonlinearity correction were
used to preprocess the original spectra. Based on the collected transmission spectra, transmitted intensity in 12 wavebands
which were the most relevant to moldy core of apples, were selected and combined. The diameter data were analyzed by
principal component analysis (PCA). The Fisher discriminate model was developed using the PCA results as independent
variables and verified. The overall accuracy rate was 93.1%, while that of the spectral model was 91.37%. In conclusion, the
multiple-factor modeling based on transmission intensity and diameter permits accurate detection of moldy core in apples.
This research can provide feasible ideas for rapid, non-destructive detection of moldy core in apples.