Camera recalibration with hand-eye robotic system

The authors address the geometric modeling aspects related to the use of a pair of video cameras mounted on a six-degree-of-freedom manipulator in a hand-eye configuration. The information obtained by these sensors will be used to generate the depth map of the environment where the manipulator works. The procedure for determining the image formation parameters is known by camera calibration, and a method to optimize these calibration parameters is presented. A calibration matrix invariant to movement is developed. The optimization process is applied to this matrix to obtain better performance in the numerical data. This optimization process is the basis for a recalibration procedure that exploits the capability of positioning of the cameras in different positions within the workspace of the robot. The first optimization process used is a recursive form of the least-squares method and the second is the Kalman filter. Experimental results obtained show that the recalibration technique gives some improvement in the numerical data and stabilizes the calibration matrix along the different positions of the vision system within the robot workspace.