A Method for Projector Calibration Encompassing the Relative Positioning with a Camera

Addressing the issues of missing world coordinate dimensional information and tedious operations in traditional projector calibration methods, this paper proposes a method for projector calibration encompassing the relative positioning with a camera. The three-dimensional information reconstructed by stereo cameras is used as the world coordinates to find the corresponding projector pixel coordinates. The projector's projection transformation matrix is solved using the singular value decomposition method. Then, the initial values of the projector's internal parameters are calculated using the properties of rotation vectors. The external parameters are determined, followed by a global optimization of all calibration parameters using a nonlinear optimization method. Experimental results show that the projector calibration algorithm is simple to operate, with a reprojection error of 0.6148 pixel.