基于近红外光谱和仿生模式识别玉米品种快速鉴别方法

文章提出了一种采用近红外光谱快速鉴别玉米品种的新方法，并对不同品种的玉米种子建立了相应的鉴别模型。对7个玉米品种共140个样本，通过近红外光谱仪扫描获得4 000～12 000 cm-1波段范围内的光谱数据。为了消除噪声、提高数据处理效率，对原始数据进行了归一化预处理，采用固定尺寸移动窗口渐进因子法(fix-sized moving window evolving factor analysis)寻找特征波段，使用主成分分析(PCA)方法得到能反映玉米种子99.96%光谱信息的5个主成分，进而利用仿生模式识别(biomimetic pattern recognition)方法建立玉米品种的鉴别模型。对于每个品种中的20个样本，随机挑选10个样本作为训练样本，其余10个样本作为第一测试集，其他品种共120个样本作为第二测试集。在对第二测试集平均正确拒识率达到99.1%的情况下，对第一测试集中的样本取得了94.3%的平均正确识别率。该方法具有较高的鉴别准确度，可以作为一种快速无损的玉米品种鉴别方法。

Fast Discrimination of Varieties of Corn Based on Near Infrared Spectra and Biomimetic Pattern Recognition

A new method for fast discrimination of varieties of corn by means of near infrared spectroscopy and biomimetic pattern recognition (BPR) was proposed and the recognition models for seven kinds of corn were built. The experiment adopted 140 samples acquired from seven varieties of corn. Firstly, a field spectroradiometer was used for collecting spectra in the wave number range of 4 000 to 12 000 cm-1. Secondly, the original spectral data were pretreated in order to eliminate noise and improve the efficiency of models, and then the characteristic spectral regions were selected by using fixed-sized moving window evolving factor analysis. Thirdly, principal component analysis (PCA) was used to compress spectral data into several variables, and the cumulate reliabilities of the first five components were more than 99.96%. Finally, according to the first five components, the recognition models were established based on BPR. For the samples in each variety, 10 samples were randomly selected as the training set. The remaining samples of the same variety were used as the first testing set, and the 120 samples of the other varieties were used as the second testing set. Under the condition that almost all the samples in the second set were correctly rejected, the average correct recognition rate was 94.3%. The experimental results demonstrated that the recognition models were effective and efficient. In short, it is feasible to discriminate varieties of corn based on near infrared spectroscopy and BPR.