论文部分内容阅读
本文讨论了模糊滤波器性能的度量问题,提出了一种简单实用的模板距离判别准则。据此准则,我们分别检查了第Ⅰ类模糊滤波器中定义的广义距离中对角距离和非对角距离的作用。结果表明,在广义距离的定义中应去掉非对角离距这一项。进而提出了一类性能好、运算最小的模糊滤波器。最后,以条码识别为例进行了计算机模拟,所得结果与理论分析基本一致。
This paper discusses the measurement of the performance of the fuzzy filter, and proposes a simple and practical template distance discrimination criterion. Based on this criterion, we examine the effects of diagonal and non-diagonal distances, respectively, in the generalized distance defined in the Class I fuzzy filter. The results show that in the definition of generalized distance should be removed off-diagonal this one. Then a kind of fuzzy filter with good performance and minimum operation is proposed. Finally, a computer simulation is carried out with barcode recognition as an example, and the results are basically consistent with the theoretical analysis.