Calibration Algorithm Based On The Stereoscopic Vision Of The Control Points Of The Permutations And Combinations

The objective of computer vision is 3D reconstruction and object recognition from computing geometry information that get by camera. The precision and dependability of camera calibration will influence the precision of computer vision. The accurate camera calibration is the important process. For cheap, little volume, light weight, easy use and getting information real-time, Camera is frequently used in computer vision. But unknown parameters and instability are disadvantages of camera. For these reasons, the precision of camera calibration can not be improved rapidly and there are most difficult and problems. So the modeling of image becomes the key in camera calibration of computer vision. In conclusion, the technique of camera calibration can sum up the traditional methods of camera calibration and the camera self-calibration.Now, most researches is focused on pursuing the more precision and speed of compute using most expensive equipment in computer vision, but no one research camera calibration used by industry purpose.Computation of DLT parameters is usually sensitive to the accuracy of control points. In order to gain more robustness without losing accuracy, a stereo calibration algorithm based on permutation and combination of control points is presented. The algorithm examines permutations and combinations of all control points for computing DLT parameter, eliminates those control points that cause aberrant perturbation, and hence obtains a set of stable control points.