基于主动视觉的相机标定方法在太阳能电池片丝印机上的应用

Application of camera calibration method based on active vision in solar cell chip screen printing machine

为了实现太阳能电池片丝印机视觉对位系统高精度标定的目标,采用线性相机模型与高精度的标定板及高精度的主动视觉平台,并采用基于平面正交运动的主动视觉标定方法以实现标定系统构建。本研究阐述了线性相机模型和相机标定方法,介绍了丝印机视觉对位系统的构成,设计了基于主动视觉标定方法的主动视觉对位平台的运动过程。标定时,令主动视觉平台按照预先设计的运动信息表动作,由相机采集每次运动后标定板上特征点的图像。根据每次平台运动采集到的特征点信息与平台运动距离之间的关系,通过采用Matlab的OPENCV算法,线性地求解了相机的内外参数,给出了相机标定的部分程序,并分析了标定误差。结果表明,采用基于平面正交运动的主动视觉标定方法,仅需主动视觉平台做有限的几次运动就能够实现相机的线性标定,实现起来比较容易,且标定精度能够满足使用需求。

In order to achieve high calibration precision in vision alignment system of the solar cell printing machine, linear camera model was adopted; high-precision calibration plate and high precision active vision platform was used in the calibration system. The active visual calibration method based on orthogonal movement was employed to achieve calibration. Camera linear model and calibration method was ex- plained. The composition of the vision alignment system of the screen printing machine was introduced ; the motion process of the alignment platform based on active vision was designed. The active vision plate was moved in accordance to a pre-designed movement information table in the calibration process, the image of the target in the calibration plate was taken by the camera after every movement. According to the re- lationship between captured target information and active vision platform movement distance during each motion of the platform, OPENCV al- gorithm in MATLAB was applied to solve the inner and outer parameters of the camera linearly, part of the calibration program was showed, and the calibration error was analyzed. And The results indicate that the calibration method base on active vision of orthogonal movement only need several movements of the active vision platform to solve the camera model linearly, it is relatively easy to implement, and able to meet the needs of calibration accuracy.