Camera Calibration Algorithm

Camera calibration is the key step of recovering 3D shape from 2D images and the first step of computer vision. It is widely used in reverse engineering, object recognition, and synthesis of virtual environment.The techniques of camera calibration can be roughly classified into three categories: self-calibration, active vision based camera calibration and traditional ways of calibration.Self-calibration doesn't need the calibration object. Only by moving the camera to take several images, it achieves calibration by using the corresponding points of two images. Since this technique is based on absolute conic and absolute quadric, the process of calibration is complex and the ability of anti-jamming is not much satisfied. The active vision based camera calibration method requires knowing exactly the movement of camera to perform the calibration. This technique needs precise equipment to catch the track of the moving camera, which is every expensive. Until now the traditional ways of calibration are the most popular technique for camera calibration. It is performed by observing a calibration object whose size or geometry is known. By taking the images of the calibration object, we can compute the corresponding relationships between the 3D points and image points, and then solve the calibration problem. Satisfactory result can be obtained by reasonable design of the calibration object.This thesis analyzes and summarizes the existing calibration techniques and mainly analyzes some traditional calibration ways, then improve the calibration method based on 1D calibration object and raise two new calibration methods: the method based on two perpendicular intersect lines and the method based on spatial triangleIn the method based on two perpendiculars intersect lines two vertical across lines are used as calibration object. Set the intersect point of the two lines to be fixed and rotate the calibration object about the fixed point freely .By taking at least 6 pictures of the...