A multilayer neural network based on multivalued neurons (MLMVN) is a neural network
with a traditional feedforward architecture. At the same time, this network has a number of …

Convolutional neural networks (CNN) have shown its excellent performance in computer
vision fields. Recently, they are successfully applied to image restoration. This paper …

Super resolution image reconstruction under unknown Gaussian blur has been a
challenging topic. Advanced optimization-based works for blind image super-resolution (SR) …

The type and extent of blur affect image quality and therefore its evaluation. This paper
presents an accurate method for blur identification and parameter estimation from one …

Motion-blurred images are usually generated when captured with a handheld or wearable
video camera, owing to rapid movement of the camera or foreground (ie, moving object …

In this paper, a criterion for objective defocus blur measurement is theoretically derived from
one image. The essential idea is to estimate the point spread function (PSF) from the line …

This letter proposes a blind image restoration method for the deblurring of remote sensing
images. A simple but robust identification method of point spread function (PSF) support is …

This paper explores an image processing method for synthesizing refocused images from a
single input photograph containing some defocus blur. First, we restore a sharp image by …

Total variation (TV) has been used as a popular and effective image prior model in
regularization-based image restoration, because of its ability to preserve edges. However …