Car Monocular Camera Self-calibration Algorithm Online

Camera calibration is a key technology of vehicle driving safety assistant system and is an essential step of extracting three-dimensional message from two-dimensional image. Its main purpose is to solve the parameters of the internal and external camera by a specific algorithm,so as to provide technical support from two-dimensional message cognize three-dimensional message. This thesis focused research the self-calibration algorithm of the on-board monocular camera.The thesis first introduces basic theoretical knowledge relative with camera calibration, such as the homogeneous coordinate, the camera imaging principle, vanishing point, absolute pomageint and so on. The chapter lays a foundation of the understanding of the camera self-calibration method.Then, the thesis realized to solve the camera parameters based on three rectangular is taken at different angles of camera self-calibration method. It requires the existence of a rectangle with side length and position unknown in the scene. Shooting for rectangular image with the camera from different angles, and then obtain a nonlinear equations with parameter r (r is aspect ratio of the rectangle) through the constraint of the camera parameter. Sloving the value of r, then the nonlinear equations changed into linear equations, the inner parameters of the camera can be obtained by solving the liner equations.In the aspect of the calibration of outer parameter, the thesis researches the relationship between the outer parameters of the camera and the lane slop in the image and voishing point first. Then, combining with the camera imaging principle, the thesis derived a simple algorithm to calculate the outer parameters of the camera.Finally, the thesis conductes a complete experimental operation combined these two parameter identification algorithmthe to calibrate the internal and external camera parameters.