Research On Camera Calibration Based On Radial Consistency Constraint Of Vanishing Point

Camera calibration technique is the key of three-dimensional information that is obtained from the two-dimensional digital images, and is also important basic computer vision research, which is applied in photogrammetry, machine vision, security monitoring, three-dimensional reconstruction, bio-medicine, remote sensing and mapping, navigation control and other fields.For the problem that the main point cannot be solved in the process of TSAI two-steps method, this work improved the calibration method of TSAI two-steps. The strict mathematical relationship between camera parameters and process parameters of TSAI two-steps method are derived and established. A camera calibration technique is presented based on the vanishing point radial direction constrained by uniformly alignment. The specific work is given as follows:1). A number of related theories about camera calibration, including the camera model, distortion model, geometric theory of vanishing point and error handling method are studied.2). The TSAI two-step calibration and single point calibration are focused on via summarizing, comparing and analyzing the existing digital camera calibration technique. The principles and concrete steps of above two algorithms are also studied. Calibration experiments are carried out to analysis the advantages.3). Vanishing point geometric constraint is improved by TSAI two-steps calibration method that is introduced in this paper for the problem that the main point cannot be solved in the calibration process of TSAI two-steps method. The strict mathematical relationship between two-vanishing points configured by camera parameters and process parameters of TSAI two-steps method are derived and established, radial distortion parameters and exterior orientation parameters of camera are solved out.4). Experiments are conducted corresponding to TSAI two-steps algorithm and the improved algorithm in this paper, and the results prove the improved algorithm has validity, high precision and stability.