二维Fisher线性判别中子模式性能的一种评价方法*

针对二维Fisher线性判别（2DFLD）方法中传统子模式选取方法计算量大、十分耗时的问题，结合影响2DFLD方法识别结果的两个主要因素——样本在投影空间的离散程度和子模式之间的相似度，提出了一种子模式性能的评价方法。首先设计子模式的构成方式；接着依据该评价方法计算各个子模式的性能指标；最后选出较优的子模式。在人耳图库、ORL人脸图库及虹膜图库上的实验结果表明，该评价方法能有效地选取较优子模式，并能够将计算时间缩短为常规子模式选择方法的近1/4，是一种有效的子模式性能评价方法。

Evaluation method for capability of sub-pattern in two-dimensional Fisher linear discriminant

In order to overcome the problem that the traditional sub-pattern choosing method in the two-dimensional fisher linear discriminant(2DFLD) algorithm cost huge computation amount and was time-consuming, a method, based on the two main factors influencing the 2DFLD recognition effect, the discreet degree of samples in projection space and the similarity between sub-patterns, for evaluating the capability of sub-pattern was contrived. Firstly, designed the form of sub-pattern. Secondly, calculated the capabilities of all kinds of sub-patterns based on this evaluation method. Finally, obtained the preferable sub-pattern. Experiments are carried out in the ear database, ORL face database and iris database which show that this evaluation method can effectively choose the preferable sub-pattern, and shorten the calculation time to about 1/4 of the conventional choosing method. It is an effective evaluation method for capability of sub-pattern.