数字摄像头路径识别技术的研究与应用  

Research and Application of Path Recognition Technology Based on Digital Camera

作　　者：胡长晖[1] 叶梦君[1] 叶天凤[2] 万里光[1] 张先鹤[1] 张佳锐[1] 

机构地区：[1]湖北师范学院机电与控制工程学院 [2]湖北理工学院电气与电子信息工程学院

出　　处：《可编程控制器与工厂自动化（PLC FA）》2012年第8期81-85,共5页Programmable controller & Factory Automation(PLC & FA)

基　　金：湖北理工学院青年项目资助(11yjz21Q)

摘　　要：本文以组态王软件和KingACT控制软件为基础,路径识别技术是自寻迹车的关键核心技术,本文介绍了基于数字摄像头OV6620的自寻迹车系统,阐述了逐点扫描法,中间向两边提取法,小范围搜索法三种提取路径导航信息的算法;详细阐述了平均值法,三点外接圆法,最小二乘拟合法和改进斜率法四种路径识别算法。通过MATLAB编程实现七种算法,输出比较各种算法优劣。实际测试表明,小范围搜索法和改进斜率法能保证高速行驶的自寻迹车具有最好的稳定性和最强的抗干扰能力。Path recognition technology is the key core technology of self-tracing vehicle system .This article introduces the self-tracing vehicle system based on digital camera OV6620.Described three kinds of extracting path navigation information methods, which are point by point scanning method, middle to both sides scanning method and scanning with small-scale method ,and elaborated four kinds of path recognition algorithms, which are average value method, three-point circumcircle method, least squares fitting method and improved slope method. Program above methods based on MATLAB. Actual tests show that scanning with small-scale method and improved slope algorithm support high-speed tracing car with the best stability and the strongest anti-jamming capability.

关 键 词：自寻迹车 数字摄像头 路径导航信息提取 路径识别算法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]