作　　者： ; ;

机构地区： 广州大学数学与信息科学学院

出　　处： 《辽宁工程技术大学学报（自然科学版）》 2010年第5期978-981,共4页

摘　　要： 为了解决三角形类型的模糊模式识别问题,以便更好地应用于生物细胞染色体形状的识别、白血球分类等实际应用问题中,采用基于给定阈值的最大隶属原则,利用合格指标因子归类法进行研究,并提出一种以三角形内角和差关系作为底数和指数变量的指数型隶属函数。结果表明:求得的隶属度确保在合格指标因子之上皆可,利用指数函数性质使所求得的隶属度差距较大、区别明显,便于识别,且与三角形三个内角紧密相关。通过数值实例,论证了该方法对染色体形状识别、白血球分类等三角形类型模糊模式识别,具有一定的参考价值和指导意义。 In order to resolve the problem associated with a triangle type fuzzy pattern recognition for the identification of chromosome in biological cells and the classification of leukocyte,a new exponential membership function is proposed,which uses the addition or subtraction of triangle angles as the base or exponent of exponential function. The function proposed is developed using the maximum membership principle which is based on the given threshold,and the acceptable index factor method. The study result shows that the membership degrees obtained are acceptable once they are above a qualified index factor. The gaps among membership degrees obtained using the exponential functions are significantly large and easily identified,and depends on the angles of triangle. A case study demonstrates the superiority of this method for resolving the triangle type fuzzy pattern recognition.

关 键 词： 模糊模式识别 三角形类型 最大隶属原则 阈值 合格指标因子 指数型隶属函数

领　　域： [理学] [理学]