Image blur often hinders radiographic data analysis. By approaching image deblurring as an ill-posed inverse problem, methods such as Tikhonov and Total Variation Regularization provide ample solutions. As non-blind methods, the regularization requires a Point Spread Function for computation and thus must be estimated. In this work, we present a comprehensive analysis of the uncertainty propagation of these processes. The findings improve the utility of regularization in image deblurring across various fields.