摘要
以山东和陕西两地产的红富士苹果作为实验对象,提出了一种不同产地苹果的分类识别法。首先对苹果的近红外光谱数据进行小波软阈值预处理,去除噪声和冗余;再采用主成分分析法(Principle ComponentAnalysis,PCA)进行降维;然后应用Fisher判决(Fisher DiscriminantAnalysis,FDA)进一步提取特征;最后使用K_近邻法进行分类识别(K_near neighbor classification,KNN)。通过实验比较,本文提出的方法能很好地实现不同产地苹果无损、快速和准确分类识别,识别正确率达到97.5%。
Taking the red Fuji apples produced in Shandong and Shanxi Provinces as the experimental objects,a classification method for identifying the apples produced in different regions is proposed.Firstly,the near infrared（NIR） spectra of the apples are preprocessed by a wavelet soft-threshold method,removing the noise and redundancy.Then,a Principal Component Analysis（PCA） method is used to reduce the dimension of the NIR data.Secondly,a Fisher Discriminant Analysis（FDA） method is used to further extract the features from the data.Finally,a K_near Neighbor Classification（KNN） method is used for the classification and identification of the apples.The experimental result shows that the proposed method can well realize the nondestructive,fast and accurate classification and identification of the apples produced in different regions.Its identification accuracy is up to 97.5%
出处
《红外》
CAS
2014年第12期41-44,共4页
Infrared
基金
四川省教育厅重点项目(12ZA070)
关键词
不同产地
近红外光谱
主成分分析
Fisher判决
K-近邻分类
different region
near infrared spectroscopy
principal component analysis
fisher decision
K_near neighbor classification
