基于半径约束的空间圆弧位姿视觉检测算法  

Visual Detection Algorithm for the Pose of Spatial Circle Based on Radius Constraint

作　　者：刘凌云[1] 罗敏[1] 吴岳敏[1] 李慧玲 马彬 

机构地区：[1]湖北汽车工业学院电气工程系,湖北十堰442002

出　　处：《组合机床与自动化加工技术》2018年第1期102-105,共4页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：湖北省自然科学基金项目(2016CFB401)

摘　　要：为了获取空间圆弧的位姿信息,建立了空间圆弧的实际影像到虚拟像机中影像的映射数学模型。以圆弧虚拟影像的拟合圆度最大为目标,提出了一种基于虚拟像机姿态参数交叉迭代、分层搜索的优化算法获取空间圆弧的姿态信息,进一步根据半径约束和几何相似原理获取空间圆弧圆心的位置坐标。仿真以及实物实验表明该方法具有很强的收敛性,可以快速且准确地获得空间圆弧的位姿信息。该方法可以推广到作业对象已知前提下的视觉位姿检测。具有一定的理论意义与工程实际应用价值。In order to obtain the pose information of spatial arc,a mathematical model about the perspective projection from spatial circle real image into virtual camera image is built. In order to detect the pose of spatial circle,an optimization algorithm aim at the maximum fitting circular arcs of virtual image is put forward firstly,based on cross-iteration and hierarchical-search of virtual camera＇s orientation parameters. Moreover,center coordinates of the spatial circle is obtained according to the principle of radius constraint and geometric similarity. The simulation and physical experiments showthat this method has a strong convergence,can obtain the pose information of spatial circle quickly and accurately. This method can be applied to the visual position detection of specific object. It has great meaning on theory and in practice.

关 键 词：空间圆弧 位姿检测 交叉迭代 分层搜索 

分 类 号：TH161[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]