Blind image restoring method based on adaptive judgment feedback balance technique

Abstract

The invention relates to a blind image restore method based on self-adapting judgment feedback balance technology. It includes the following steps: gaining part restoring image through forward direction filter, inputting the iteration result of NL filter to feedback filter to gain part degeneration factor, taking estimation and projection, calculating the output difference of NL filter and image estimating to generate the iteration error to every pixel, and gaining total error of the blind image restoring, updating parameter of forward direction filter and feedback filter and repeating the previous process until the total error is less than the reference threshold value, and output image. The advantages of the invention are simple arithmetic structure, good astringency, low computing complex, and good image restoring effect.

Description

technical field [0001] The invention belongs to the field of blind image restoration in computer image processing, in particular to a blind image restoration method based on adaptive decision feedback equalization technology. Background technique [0002] The purpose of image restoration is to process the distorted image (here we call it a degraded image), so that it tends to be restored to a real image without degradation. If there is enough known information about the degradation process, the degradation process can be modeled, and The real image can be restored by using the opposite process. However, in many practical situations, the image degradation process is usually unknown, and the information of the real image is rarely known. Therefore, the real image must be used in some or Determined directly from observed images without any degradation process and real image information, this restoration process is called blind image restoration. Blind image restoration has a w...

AI Technical Summary

Problems solved by technology

The existing blind image restoration usually adopts iterative blind deconvolution method (IBD), simulated annealing method (SA), autoregressive moving average (ARMA) parameter estimation method, non-negative support domain recursive inverse filtering (NAS-RIF) method, However, these methods have defects: among them, the algorithm of the IBD method lacks reliability, that is, the convergence and uniqueness of the algorithm are uncertain; in addition, the image restoration is very sensitive to the initial estimate, and there will often be unstable situations; the disadvantages of the SA method It is too slow to repeatedly converge to the global minimum of the cost function, and the convergence speed largely depends on the decreasing speed of the probability

The disadvantage of the ARMA parameter estimation method is that as the order of the parameter model increases, the number of unknown parameters will increase, the number of optimal solutions will also increase, and the number of initial points required for the algorithm to converge to the global minimum point will increase so much that Unacceptable; the disadvantage of the NAS-RIF method is that the constructed inverse filter has a high-pass property, so high-frequency noise must be amplified, so the restoration effect on images with low signal-to-noise ratio (SNR) is not good

Images