The traditional projector calibration always assumes that the reprojection errors on the projector image are independent and identically distributed and utilize the same method as the camera calibration. Actually, even if the measured points on the camera image are independent and identically distributed Gaussian variables, it is impossible to obtain the statistical characteristics of the reprojection errors on the projector image, because the relation between the original measurement error and the reprojection error is nonlinear. We propose a method to estimate the projector parameters according to the camera image reprojection error. It does not require us to know the statistical characteristics of the projector image point and makes full use of the background knowledge of the camera image noise, because our cost function does not concern the reprojection error on the projector image as is the case in the traditional method. The simulations and experiments affirm that this method has higher precision, even if the real noise is not normal.