基于可见/近红外透射光谱结合蜜獾算法优化支持向量机的糖心苹果鉴别

Classification of Watercore Apples Based on Visible/Near Infrared Transmission Spectroscopy Combined with Honey Badger Algorithm-Support Vector Machine

传统的糖心苹果鉴别方法具有破坏性和不可逆性,无法大范围推广使用。为实现糖心苹果和健康苹果的快速准确分类,提出基于可见/近红外透射光谱结合蜜獾算法优化支持向量机的糖心苹果鉴别方法。首先采集正常苹果和疑似糖心苹果样本的3个不同方向的可见/近红外透射光谱,并利用最大最小值归一化对原始光谱数据进行预处理,然后使用主成分分析对预处理完的数据进行降维和特征提取,再取前10个主成分作为降维后的样本数据,最后将降维后的样本数据输入支持向量机进行分类,结果发现分类效果一般。引入蜜獾算法对支持向量机进行优化,建立新模型,通过结果表明,方向二为光谱数据采集的最佳方向,新模型可以实现对健康苹果和糖心苹果的快速准确分类,为糖心苹果的鉴别和其他果蔬的分类提供新思路。

The traditional watercore apple identification method is destructive and irreversible,and cannot be widely used.In order to realize the rapid and accurate classification of watercore apples and healthy apples,a watercore apple identification method based on visible/near-infrared transmission spectroscopy combined with honey badger algorithm to optimize support vector machine was proposed.Firstly,the visible/near-infrared transmission spectra of normal apples and suspected watercore apple samples in three different directions were collected,and the original spectral data were preprocessed by maximum and minimum normalization.Secondly,principal component analysis was used to reduce the dimension and extract the features of the preprocessed data.Then,the first ten principal components were taken as the sample data after dimension reduction.Finally,the sample data after dimension reduction were input into support vector machine for classification,and the classification results were general.The honey badger algorithm is introduced to optimize the support vector machine and a new model is established.The results show that the direction two is the best direction for spectral data acquisition.The new model can realize the rapid and accurate classification of healthy apples and watercore apples,provide new ideas for the identification and classification of other fruits and vegetables.