A non-iterative blind image deblurring algorithm based on OTF estimation

This paper proposes a non-iterative algorithm on blind image deblurring. This algorithm can restore the degraded images which are blurred by class G. This algorithm is based on that most images spectral amplitude have the similar power law distribution. In accordance with the power law distribution of natural image spectrum, a curve model is proposed to approximate the spectrum of the true image. The OTF(optical transfer function) is estimated by comparing the spectrum of the degraded image with the reconstructed one. Then, the image is restored by employing the estimated OTF and the Wiener filtering. The experiments show that this algorithm obtains a more accurate OTF, and this algorithm can reduce ringing artifacts as compared with some existing algorithms. The quality of the restored images has been enhanced significantly.