A Hybrid Regularization-Based Multi-Frame Super-Resolution Using Bayesian Framework

Abstract

The prime purpose for the image reconstruction of a multi-frame superresolution is to reconstruct a higher-resolution image through incorporating the knowledge obtained from a series of relevant low-resolution images, which is useful in numerous fields. Nevertheless, super-resolution image reconstruction methods are usually damaged by undesirable restorative artifacts, which include blurring distortion, noises, and stair-casing effects. Consequently, it is always challenging to achieve balancing between image smoothness and preservation of the edges inside the image. In this research work, we seek to increase the effectiveness of multi-frame super-resolution image reconstruction by increasing the visual information and improving the automated machine perception, which improves human analysis and interpretation processes. Accordingly, we propose a new approach to the image reconstruction of multi-frame super-resolution, so that it is created through the use of the regularization framework. In the proposed approach, the bilateral edge preserving and bilateral total variation regularizations are employed to approximate a high-resolution image generated from a sequence of corresponding images with low-resolution to protect significant features of an image, including sharp image edges and texture details while preventing artifacts. The experimental results of the synthesized image demonstrate that the new proposed approach has improved efficacy both visually and numerically more than other approaches.

To shift or to rotate? Comparison of acquisition strategies for multi-slice super-resolution magnetic resonance imaging

“…However, since MS-SRR estimation involves solving an ill-conditioned inverse problem, unregularized estimators, although unbiased, often have an unacceptably high variance. Furthermore, when unregularized estimators are used, small modeling errors may lead to reconstruction artifacts that prevent the MS-SRR image's diagnostic use (Khattab et al, 2020). These artifacts can be minimized by regularizing the MS-SRR problem, as qualitatively shown by Shilling et al (2008).…”

Compressive Sensing in Image/Video Compression: Sampling, Coding, Reconstruction, and Codec Optimization

“…The downsampling, as observed in approaches like Bilinear and Bicubic, serves as a pre-processing step, while up-sampling methods are employed as a post-processing step. This type of framework usually achieves a superior rate-distortion while utilizing network-based and high-quality super-resolution algorithms [97][98][99]. However, there are few studies that conduct this idea into CS for codec optimization.…”

AFBNet: A Lightweight Adaptive Feature Fusion Module for Super-Resolution Algorithms

“…These gaps are filled with zero, so the feature map size can be calculated given the step size, as shown in Eq. (9).…”

“…It has significantly contributed to the advancement of various fields, including public safety [1,2], medical imaging [3], target detection [4], satellite remote sensing [5][6][7], and new media [8]. Single Image Super Resolution (SISR) algorithm can reconstruct high-resolution images from lower ones, which can recover more image details and bring better visual effects [9][10][11][12]. Dong et al introduced a Super-Resolution Convolutional Neural Network (SRCNN) [13] as the first super-resolution convolutional neural network.…”