协同进化算法在散斑条纹角点识别中的应用  

APPLICATION OF CO-EVOLUTIONARY ALGORITHM IN SPECKLE STRIPE CORNER RECOGNITION

作　　者：李学哲[1] 孙文卿 崔国增 Li Xuezhe;Sun Wenqing;Cui Guozeng(School of Electronic and Information Engineering,Suzhou University of Science and Technology,Suzhou 215009,Jiangsu,China;School of Math and Physics,Suzhou University of Science and Technology,Suzhou 215009,Jiangsu,China)

机构地区：[1]苏州科技大学电子与信息工程学院,江苏苏州215009 [2]苏州科技大学数理学院,江苏苏州215009

出　　处：《计算机应用与软件》2019年第10期233-237,250,共6页Computer Applications and Software

基　　金：江苏省苏州市科技计划项目(SYG201428)

摘　　要：介绍激光散斑干涉位移测量原理以及基于该测量方法的重定位系统,提出用于识别条纹信息的协同进化算法。采用线性降维法大幅降低搜索难度,并详细研究针对条纹角点协同算法的亲和度计算、抗体浓度控制和克隆选择、交叉及变异操作等方法。该算法可实时精确提取条纹中心线,并识别条纹数、条距和条角信息。研究表明将上述测量信息应用于移动控制平台,能够满足高精度重定位要求。This paper introduced the principle of laser speckle interferometry displacement measurement and the relocation system based on this measurement method,and proposed a co-evolutionary algorithm for recognition stripe information.Linear dimensionality reduction method was used to reduce the difficulty of searching.Affinity calculation,antibody concentration control and clone selection,crossover and mutation operation methods for stripe corner collaboration algorithm were studied in detail.This method could extract the stripe center line accurately in real time,and recognize the stripe number,distance and angle information.The research shows that the application of the above measurement information to the mobile control platform can meet the requirements of high precision relocation.

关 键 词：协同进化 降维 散斑条纹 浓度抑制 条纹角点 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]