提出了基于特征矩阵和改进的奇异值分解的人耳识别算法。采用特征矩阵及Fisher 最优鉴别分析方法将原始样本向最优鉴别矢量投影,有效地降低了维数,再进行奇异值分解后所得到的正交矩阵,进行人耳识别,实验结果验证该方法在人耳识别应用中的有效性。 关键词:人耳识别,广义Fishier 鉴别分析,特征矩阵,奇异值分解 Abstract: An algorithm based on eign matrix and the improved SVD is proposed for ear recognition. Eign matrix and the generalized optimal discriminant analysis is used for decreasing the dimensions of images. The results of experiment demonstrate that the algorithm is effective and quickly for ear recognition. Key words: Ear recognition, Generalized Fisher discriminant analysis, eign matrix, SVD