广角镜头的摄像机非线性标定技术

Technique for Wide-angle-lens Camera Calibration

在路面检测的计算机视觉系统中,采用了线结构光照明和摄像机采集外界三维信息。为了实现在大范围内的测量,摄像机采用了广角镜头摄像,同时也引入了比较大的图像畸变。为了精确标定摄像机,采用比较接近实际情况的几何模型,并采用非线性迭代解方程的方法对摄像机进行标定,求解摄像机的内外参数。实验精度可以达到0.1mm。该方法简洁实用,可以采用这种方法在实际当中对系统进行标定。

In the computer vision of road measurement,line-structured light illumination and 3D information collected by camera are used. To conduct a large scale measurement,wide-angle lens is used in the camera and at the same time image distortion is introduced. In order to have the camera accurately calibrated,a geometric model close to real situation is deployed and nonlinear iterative equation is used to calibrate the camera for deducing its internal and external parameters. The experimental precision can be as high as 0.1mm. the method is simple and practical for calibrating the system.