The key point of super-resolution(SR) reconstruction based on maximum a posterori(MAP) is the choice of
the regularization and it is achieve in frequency domain mostly. Therefore, a self-adapting technology for image super-
resolution reconstruction in frequency domain and time domain is proposed. Firstly, according to the characteristics of
different images, frequency regularization(FR) and time regularization(TR) are defined. Then, an image reconstruction
model is given. The frequency domain and time domain weights are introduced to strengthen the algorithm more adaptive.
Finally, iterative scheme are developed to get the more accurate image for SR reconstruction according to conjugate gradient.
The experimental results on both real and synthetic data show the effectiveness of the proposed algorithm.