Research On The Methods Of Camera Self-Calibration

Camera calibration is a necessary step for obtaining the 3D information from the 2D images, which has become a hotspot in the fields of computer vision. Recently, in ICCV, ECCV, CVPR, ICPR and some important international magazine, many papers of this area were published. This dissertation analyzed some problems in the area of self-calibration and studied some existing self-calibration algorithms, the main works of which are as follows:1. Analyzed the present research status of camera calibration and introduced the basic principles for camera calibration.2. Analyzed the typical camera calibration based on the circular points. Improved the calibration method based on planar template. The experimental data prove the validity of this method.3. Proposed a new calibration method. In the new method, the unknown scale factors in Kruppa equation are estimated first, and then they are optimized by simulated annealing algorithm, and the camera is calibrated by linear method finally. The experimental results show that the method is quite effective.