| Camera calibration is an important technology in the vision measurement field, a main factor determining the accuracy of vision measurement. Considering the demand for the high accuracy of measurement in the exist built measurement system for Coding structure light. The paper did some research on the problems, such as the methods of camera calibration, the evaluation for the accuracy and corner detector for camera calibration, and the camera control. Here we take a projector as an example for a camera and study its method of camera calibration.1. Analysis the accuracy of step approach, Absolute conic method and Plane Method, and Present an evaluation method of the accuracy for the camera calibration. First, get the coordinates for the Adjacent two lattice points in the check board and the distance between the points based on the camera calibration parameters, then make a contrast with the actual distance. Next test and verify the calibration method and the evaluation method of the accuracy for the camera calibration, and choose the Plane Method as the calibration method in this measurement system.2. It is important to detect the image coordinates of corner detector for camera calibration accurately to ensure the accuracy of the calibration. In the detector, we first binaries the gray image using adaptive threshold value method to reduce the effect of edge fuzzy for comer localization, and extract a certain amount of feature points using the Harris detector, and then process the detected corner using Susan detector, finally remove the false corner due to feature point. The method helps to improve the accuracy of camera calibration.3. Program the code for the camera control on the platform VC, and accomplish the integrations with a camera, a projector and a set of personal computer..4. Conduct the experiment for camera calibration and the evaluation for the accuracy of the camera calibration, and analysis the results to offer grounds for choosing proper calibration method. |
PDF Full Text Request