In order to eliminate the interference of out-of-focus optical images in in-focus image formed within a diffractive optic imaging spectrometer, a regularization method is proposed that can be applied to restore its spectral image. Through introduction of a stable functional and addition of some priori or assumed conditions, the inverse filter function is regularized so as to reduce the perturbance in image restoration process from the systematic noise and transfer the recovering process into the problem to solve a well-posed equation. By adjusting and choosing appropriate regularization parameter, our restored spectral images are the best and are evaluated with their Mean Square Error (MSE) and absolute error distribution. Computer simulation experiment shows that regularization algorithm is effective for restoring the spectral images formed with diffractive optic imaging spectrometer and suppressing its noise’s influence on the restored images.