三角形识别方法新探
首发时间:2019-05-10
摘要:为了解决传统三角形识别以及其他人的改进方法中由于隶属函数的隶属度整体连续性造成的待识别三角形明明不符合各类三角形的隶属条件,却空有"虚假"隶属度,与实际完全不符的情况以及当一个三角形对多个类型的模糊模式均具有较高隶属度情况下依靠最大隶属度原则所得结果与人们实际意愿倾向不符的情况。在此,将三角形的识别分类、分层,并对隶属函数进行分段,通过判别其是否符合模糊范围,来确定其是否具有模糊性,进而消除不合理的隶属度。把各类三角形看作符合特定条件的特例,由繁到简,依次对正三角形、等腰直角三角形,等腰三角形,直角三角形与非典型三角形进行逐层筛选,确定其模糊范围并结合最大隶属度原则确定待识别三角形种类。结果表明:改进方法所得的识别更加符合常理,有效解决了"虚假隶属度"问题,对于各种简单模式的模糊识别,具有一定的参考价值和指导意义。
关键词: 分层 分段隶属函数;最大隶属度;模糊范围;模糊性
For information in English, please click here
A New Approach to Triangle Recognition
Abstract:In order to solve the problem of traditional triangle recognition and other improved methods, because of the continuity of membership degree of membership function, the triangle to be identified clearly does not meet the membership conditions of all kinds of triangles, but there is no "false" membership degree, which is totally inconsistent with the actual situation.In this paper, triangles are classified and stratified, and membership functions are segmented to determine whether they conform to the ambiguity range, and then to eliminate unreasonable membership.Regarding all kinds of triangles as special cases that meet specific conditions, from complexity to simplicity, we screen regular triangles, isosceles right triangles, isosceles triangles, right triangles and atypical triangles step by step, determine their ambiguity range and determine the types of triangles to be identified by combining the principle of maximum membership degree.The results show that the membership degree of all kinds of triangles obtained by the improved method conforms to common sense, and effectively solves the problem of "false membership degree", which has certain reference value and guiding significance for the fuzzy recognition of various patterns.
Keywords: Hierarchical and piecewise membership functions Maximum membership Fuzzy range Fuzziness
基金:
引用
No.****
动态公开评议
共计0人参与
勘误表
三角形识别方法新探
评论
全部评论0/1000