Vision calibration for automated guided vehicle based on static and motion two states

A vision calibration technique based on two states of static and motion was proposed for a vision-based automated guided vehicle. Firstly, intrinsic parameters, radial distortion parameters and external parameters were estimated using planar template in static scene, and then union model of correcting distortions was designed for three individual image distortion models. Finally, rotation and translation parameters between corrected image and automated guided vehicle coordinate system can be calibrated in two motion cases. Experimental results show that the technique has the features of good flexibility and high precision.