线结构光参数的简易标定方法  

A simple method to calibrate parameters of the line structured light

作　　者：陈新禹[1] 马孜[1] 汪洋[1] 黄进[1] 陈天飞[1] 

机构地区：[1]大连海事大学自动化研究中心,辽宁大连116026

出　　处：《光电子．激光》2013年第3期563-568,共6页Journal of Optoelectronics·Laser

基　　金：国家科技重大专项(2009ZX04001-021)资助项目

摘　　要：为了实现高效率、低成本、高精度地标定线结构光的参数,本文提出了一种基于单个平面同心面靶标标定线结构光参数的新方法。该方法基于三点透视模型和交比不变原理计算光平面上标定点的三维坐标,利用最小二乘法拟合平面方程。在摄像机视野范围内自由移动平面靶标,同时保持结构光投射到靶标上同心圆之外的区域,采集至少两幅图像即可完成线结构光参数的标定。实验结果表明:光平面标定的均方根误差约为0.05mm,传感器测量精度的相对误差约为0.3%,系统整体测量精度约为0.1mm,说明本文标定方法具有较高的标定精度。并且该方法同时适用于其它类型结构光的标定,通用性较强,具有获取标定点效率高、计算简单且操作灵活方便等优点,适合现场标定。In order to calibrate parameters of the line structured light with high efficiency, low cost and high precision,a calibration technique based on planar target with only a pair of concentric circles is proposed in this paper. The method determines 3-D coordinates of points on the line structured light using the model of perspective-three-point （P3P） and the principle of cross-ratio invariance, and then the equation of the plane of the line structured light is obtained from the points by least square method （LSM）. The planar target is moved freely within the field of the camera＇s view while keeping the structured light projecting onto the outside region of concentric circle on the target ,and at least two images are acquired to complete the calibration. Finally, the calibration accuracy and overall scanning residuals are analyzed through some experiments. Experimental results indicate that this calibration technique has high accuracy,the RMS error of structured light calibration is approximately 0. 05 mm, the relative error of the measurement accuracy is about 0. 3 % and the overall measurement accuracy is about 0. 1 mm. Furthermore,the method we proposed is also suitable for other types of structured light with high generality. It concludes that the process is efficient, the calculation is simple, and the whole operation is easy, flexible and suitable for field calibration.

关 键 词：视觉测量 线结构光 标定 三点透视模型(P3P) 交比不变 

分 类 号：TP212.6[自动化与计算机技术—检测技术与自动化装置] TP242.62[自动化与计算机技术—控制科学与工程]