基于字形编码与拼音编码的近似商标辨识算法研究  

Research on Approximate Trademark Identification Algorithm Based on Glyph and Pinyin Coding

作　　者：祁俊辉 龙华[1] 赖华[1] 毕丹宏[2] QI dun hui;LONG Hua;LAI Hua;BI Dan hong(Faculty of Information Engineering and Automation,Kunming University of Science and Technology,Kunming 650000;Jiangsu Posts and Telecommunications Planning and Designing Institute CO,LTD,Nanjing 210000,China)

机构地区：[1]昆明理工大学信息工程与自动化学院,云南昆明650000 [2]江苏省邮电规划设计院有限责任公司,江苏南京210000

出　　处：《软件导刊》2018年第6期77-80,84,共5页Software Guide

摘　　要：为了实现高效、快速的商标自动评审工作,提出将商标文字通过字形编码、拼音编码等方式映射为数字型字符串和字母型字符串的辨识算法。算法依照字形和拼音描述商标信息,配合改进后的Jaro-Winkler distance算法计算出商标文字的字形相似度和拼音相似度,并根据相似匹配算法得出商标文字的相似程度。实验结果表明,该方法所得结果有更好的辨识率,符合常规人体视觉和逻辑判断。In order to realize efficient and rapid automatic evaluation of trademark,an identification algorithm is proposed to map the trademark text into numeric and alphabetic string using ideographic and Pinyin encoding.The algorithm is used to describe the trademark according to the font and Pinyin,and with the improved Jaro-Winkler distance algorithm the similarity of trademarks font and Pinyin coordinated is calculated.The similarity of words in trademark is calculated by similarity-matching algorithm.The experimental results show that the method conforms to judgment of physical vision and logic and has high recognition rate.

关 键 词：商标辨识算法 近似商标 形近字 近音字 相似度 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]