Blurred Image Restoration Algorithms Based On Neural Network

Restoration of degraded image is an important and widely studied problem in digital image processing. Image restoration is a process of removal or minimization of degradation from an image distorted by blurring and additive noise. The purpose of restoration is to process the blurred image to make it look clear.Various image restoration methods have been proposed, like inverse filter, Wiener filter, maximum entropy, iterative blind deconvolution, etc. However, these traditional methods are deficient in solving the problem of function approximation. Artificial neural network has its unique advantages in this aspect, for it is a type of large-scale nonlinear dynamical system, characteristic of high-speed parallelism calculation, great robustness, strong capacity of self-adaptive, self-organization and self-learning. BP neural network has the ability to learn the mapping relationship between input and output in an implicit way. In other words, there is no need to express the relationship explicitly. So it is suitable for the situation in which the Point Spread Function is difficult or impossible to know.The main contributions of this paper are:1. This paper proposed a new method for blind image restoration using a BP neural network. A sliding-window based technique is applied to obtain the features of the blurred image for dimension reduction. A...