首先计算数据的类内和类间散度矩阵,得到差形式的目标函数;然后进行特征值分解,得到映射矩阵;最后利用实际植物叶片数据集进行植物分类试验.结果表明,该算法对植物分类是有效可行的.%At first, the within-scatter and between-scatter matrices are calculated respectively, and the ' minus' form objective function is obtained. Secondly, the projection matrix is obtained by calculating the eigenvector. Finally, the plant taxonomy experiment if carriedout based on real plant leaf database. Results show that the proposed method is effective and feasible for plant taxonomy
展开▼