单目式自主机器人视觉导航中的测距研究  

Distance Measurement in Visual Navigation of Monocular Autonomous Robots

作　　者：吴刚[1,2] 唐振民[1] 

机构地区：[1]南京理工大学计算机科学与技术学院,江苏南京210094 [2]南京工程学院车辆工程系,江苏南京211167

出　　处：《机器人》2010年第6期828-832,共5页Robot

基　　金：国家自然科学基金资助项目(90820306);总装预研基金资助项目(9140A16070106BQ02)

摘　　要：针对传统自主式机器人视觉测距方法的缺陷,提出了一种建立在B对偶空间几何学上的单目摄像机标定方法,并将其应用到单目式自主机器人的视觉测距中,与OpenCV标定方法相比,新的标定方法对CCD摄像机内参数的计算更为准确,并减小单目视觉测距的总体误差.同时,在引入运动目标自动提取算法的基础上,将目标车辆的自动检测与实时测距融合为一个完整的系统,通过实际的测量及对比试验,验证了所提出方法的有效性.Aiming at the drawbacks in traditional visual distance-measurements for autonomous robots,a monocular camera calibration technique based on B-dual-spacial geometry is brought forward and applied to visual distance-measurement of monocular autonomous robots.Compared with OpenCV＇s calibration technique,the new calibration method can compute camera＇s inner parameters more veraciously,and it can obviously reduce the total error in monocular visual distance-measurement. Simultaneously,the method combines real-time distance measurement with automatic detection for objective vehicles by incorporating the automatic extration algorithm of moving objects.The validity of the presented method is verified by the actual measurements and contrastive experimentations.

关 键 词：视觉导航 单目 测距 摄像机内参数 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]