一种基于平面标靶的线结构光视觉传感器标定方法  

A calibration method for line structured-light vision sensor based on a plane target

作　　者：吴庆华[1,2,3] 何涛[2,3] 史铁林[1] 

机构地区：[1]华中科技大学机械科学与工程学院,湖北武汉430074 [2]湖北工业大学机械工程学院,湖北武汉430068 [3]湖北工业大学先进制造与质量工程湖北省重点实验室,湖北武汉430068

出　　处：《光电子．激光》2013年第2期297-301,共5页Journal of Optoelectronics·Laser

基　　金：国家自然科学基金(51275158);湖北省教育厅产学研合作重点(C2010029)资助项目

摘　　要：提出了一种适合于现场的线结构光视觉传感器标定方法。建立了标定的数学模型,设计了一种平面点阵标靶,提出了坐标映射方法。根据结构光条纹特征点的图像坐标和对应在标靶坐标系下的坐标,以及相机内参,计算出标靶坐标系到摄像机坐标系的转移矩阵,再由转移矩阵得到特征点在摄像机坐标系下的坐标;在视场范围内,平面标靶按不同位姿摆放多次,获取投射在标靶上的所有特征点,对这些特征点进行平面拟合,得到结构光平面在摄像机坐标系下的方程。对标定的精度进行了验证,实验表明,本文方法标定过程简单,精度较高,适合于结构光传感器的现场标定。A field calibration method for the line structured,light vision sensor is put forward. A plane tar- get with solid circle array is designed. A calibration mathematical model is established, and a coordinate mapping approach is proposed. Based on the coordinates of the feature points in image coordinate sys- tem, the corresponding coordinates in target coordinate system,and the intrinsic parameters of the cam- era, a transformation matrix from target to camera coordinate system is estimated. Using the transforma- tion matrix, the coordinates of the feature points in camera coordination system are calculated. The target is freely put vc^thin the field of vision at different positions and postures several times, and more feature points are extracted. In camera coordinate system,all feature points are fitted to an ideal plane. And the equation of the line structured,light plane is obtained in the camera coordinate system. The experiment is performed to test the precision of the calibration method. The result shows that this calibration process is simple,and the calibration precision is high. It can be used to calibrate the line structured,light vision sensor on spot.

关 键 词：线结构光 平面标靶 现场标定 坐标映射 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]