High performance super-resolution reconstruction of multi-frame degraded images with local weighted anisotropy and successive regularization

Gao, Ming; Qin, Shiyin

doi:10.1016/j.ijleo.2015.08.119

Cited by 10 publications

(3 citation statements)

References 38 publications

(58 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We select the first thirteen LR images through each video and we use the diffusion registration approach [36] to estimate the motion between the LR images for all the restored methods including our model. The recovered results are measured to the above relevant approaches, such as: anisotropy local weighted SR (LWASR) [24], Bregman algorithm through morphologic prior function (BIMR) [49], total fractional-order variation model (TFOM) [58], TV regularization with Bregman iteration [44], Bregman distance SR method (IMSR) [38] and features preserving high-order PDE (EPDE) [37]. The super-resolved images with a decimation factor of r = 4 using the above SR methods and compared to our model are depicted in figures 15 and 16.…”

Section: Video Sequencesmentioning

confidence: 99%

Bilevel optimal parameter learning for a high-order nonlocal multiframe super-resolution problem

Laghrib,

Zahra Ait Bella,

Nachaoui

et al. 2023

Inverse Problems

View full text Add to dashboard Cite

This work elaborated an improved method to multiframe super-resolution (SR), which involves a nonlocal first-order regularization combined with a nonlocal p-Laplacian term. The nonlocal TV term excels at edge preserving, whilst the nonlocal p-Laplacian is commonly used to perfectly reconstruct image textures. Firstly, we discuss the existence and uniqueness of a solution to our new model in a well posed framework. Then, we derive a modified Primal-dual iteration to compute the super-resolved solution. Furthermore, we introduce a new bilevel optimization approach to learn two regularization parameters. The included tests validate that the introduced optimization procedure performs favorably compared to numerous SR approaches in terms of efficiency and accuracy.

show abstract

Section: Video Sequencesmentioning

confidence: 99%

Bilevel optimal parameter learning for a high-order nonlocal multiframe super-resolution problem

Laghrib,

Zahra Ait Bella,

Nachaoui

et al. 2023

Inverse Problems

View full text Add to dashboard Cite

show abstract

“…The used regularization is inspired from the TV and morphological regularizations, respectively. To increase the performance of the SR reconstruction against high intensity of noise, two new regularization terms were motivated in the SR context [16]. This later is based on a local variable anisotropy regularization with convenient Bregman iteration optimization, which can pontifically reduce the noise with some blur apparition.…”

mentioning

confidence: 99%

Learning nonlocal weights for second-order nonlocal super-resolution

Laghrib,

Bella,

Nachaoui

et al. 2025

DCDS-S

View full text Add to dashboard Cite

This research introduces an enhanced approach for multiframe super-resolution (SR) that incorporates a bilevel optimization technique for learning the space variable weights parameter in the second-order nonlocal term. This novel approach effectively preserves image edges and ensures accurate reconstruction of image textures. The main contributions of this work include the establishment of the existence and uniqueness of the solution within a well-posed framework and a bilevel optimization procedure to compute the weights α. The experimental results demonstrate the efficiency and accuracy of the proposed method compared to existing super-resolution approaches. These findings highlight the practical effectiveness and potential of the proposed approach in real-world applications.

show abstract

“…Other relevant SR approaches were proposed in [48], where the authors used the Bregman iteration for rapid SR image reconstruction with TV regularization and morphological regularization respectively. In the same principle, two new types of regularization terms were introduced in the SR context [19], based on a local weighted anisotropy regularization and successive regularization toward iteration process using the Bregman iteration, which reduces efficiently noise but not tackle the blur effect. A more recent SR approach based on Huber-Norm using Bregman distances was proposed in [32] with more consistency against contrast loss while strong edges and contours are well preserved with less blur.…”

mentioning

confidence: 99%

A nonlinear fourth-order PDE for multi-frame image super-resolution enhancement

Laghrib¹,

Chakib²,

Hadri³

et al. 2020

Discrete &Amp; Continuous Dynamical Systems - B

View full text Add to dashboard Cite

The multiframe super-resolution (SR) techniques are considered as one of the active research fields. More precisely, the construction of the desired high resolution (HR) image with less artifacts in the SR models, which are always ill-posed problems, requires a great care. In this paper, we propose a new fourth-order equation based on a diffusive tensor that takes the benefit from the diffusion model of Perona-Malik in the flat regions and the Weickert model near boundaries with a high diffusion order. As a result, the proposed SR approach can efficiently preserve image features such as corner and texture much better with less blur near edges. The existence and uniqueness of the proposed partial differential equation (PDE) are also demonstrated in an appropriate functional space. Finally, the given experimental results show the effectiveness of the proposed PDE compared to some competitive methods in both visually and quantitatively.

show abstract

High performance super-resolution reconstruction of multi-frame degraded images with local weighted anisotropy and successive regularization

Cited by 10 publications

References 38 publications

Bilevel optimal parameter learning for a high-order nonlocal multiframe super-resolution problem

Bilevel optimal parameter learning for a high-order nonlocal multiframe super-resolution problem

Learning nonlocal weights for second-order nonlocal super-resolution

A nonlinear fourth-order PDE for multi-frame image super-resolution enhancement

Contact Info

Product

Resources

About