Minimum distance classifier based on between-class and within-class scatter is proposed.
提出了一种基于类内类间离散度的最小距离分类器设计方法。
The experimental results show that the KDDA features achieve the best recognition performance, and the SVM classifier outperforms the minimum Euclidean distance classifier.
结果表明,SVM分类器的识别性能优于最小欧氏距离分类器,且KDDA特征的识别性能最优。
应用推荐