基于NLP的PS图像模板匹配智能识别算法  

Intelligent recognition algorithm of PS image template matching based on NLP

作　　者：苏圆婷 SU Yuan-ting(Ya’an Polytechnic College,Ya’an 625000,Sichuan Province,China)

机构地区：[1]雅安职业技术学院,四川雅安625000

出　　处：《信息技术》2024年第4期161-165,共5页Information Technology

摘　　要：在PS图像模板匹配识别研究中,针对信息丢失严重,自身的全局运算能力需要进一步提高的问题,提出基于NLP的PS图像模板匹配智能识别算法。利用自然语言处理技术描述待识别目标方向特征,对目标方向码进行二进制压缩映射处理,同时应用窗口重叠机制,避免偏移现象出现,将压缩映射后的方向特征和输入的方向特征作匹配处理,实现对目标的精准识别。实验结果表明:设计的基于自然语言处理技术的智能识别算法在处理纹理特征比较少的目标时,识别信息较完整,能够精准匹配识别目标区域,全局运算能力得到了增强。In the research of PS image template matching recognition,to tackle the serious loss of information and the need to further improve its global computing ability,an intelligent recognition algorithm of PS image template matching based on NLP is proposed.The natural language processing technology is used to describe the direction features of the target to be identified,and the binary compression mapping is carried out for the target direction code.At the same time,the window overlap mechanism is applied to avoid the offset phenomenon.The direction features after compression mapping and the input direction features are matched to realize the accurate identification of the target.The experiment results show that the when the intelligent recognition algorithm based on natural language processing technology dealing with the targets with few texture features,the identification information is complete,which can accurately match and identify the target area,and the global computing ability is enhanced.

关 键 词：自然语言处理 图像模板 智能识别 自动匹配 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]