In video-based pipeline inspection, it is essential that the inner wall image be reconstructed from panoramic image without distortion, and this can be done only after the camera be calibrated. For simplified systems without conical reflector, based on the pinhole imaging model, how to reconstruct a distortion-less unrolling image from a panoramic image of pipeline inner wall with only one calibration parameter is analyzed, and a new camera calibration method is proposed. The new calibration method makes use of two images of a circle that are taken at different distances, and the parameter is obtained by measuring its physical diameter manually and calculating its image diameters through curve fitting automatically. Experiment demonstrates that this calibration method is simple and with high precision.%在视频管道检测系统中重建管道内壁图像时必须解决图像畸变与摄像机标定问题.针对无锥面反射镜的简化系统,在针孔成像模型下,分析了如何利用一个标定参数将全景图像重建成无畸变的内壁展开图像.提出一种新的摄像机标定方法,利用已知直径的圆在两个不同位置拍摄图像,通过曲线拟合求出其像的直径,进而得到该标定参数.实验证明了这种标定方法具有操作简单、精度高等特点.
展开▼
