MF-LRTC: Multi-filters guided low-rank tensor coding for image restoration

Lu, Hongyang; Li, Sanqian; Liu, Qiegen; Wang, Yuhao

doi:10.1109/icip.2017.8296653

Cited by 6 publications

(3 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(3) Nonlinear activation layer In order to solve the problem of inseparable nonlinearity, the convolutional neural network usually introduces an activation function layer, so that the network has the learning ability of nonlinear mapping. Common activation functions include Sigmoid function, ELU function, and ReLU function [17][18][19]. The formulas are as follows:…”

Section: Image Restoration Processing Of Old Photosmentioning

confidence: 99%

An Old Photo Image Restoration Processing Based on Deep Neural Network Structure

Wang

2022

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

Old photos retain precious historical image information, but today’s existing old photos often have varying degrees of damage. Although these old photos can be digitally processed and then restored, the restoration of old photos involves multiple areas of image restoration and has various types of degradation. Currently, there is no unified model for repairing multiple types of degradation of old photos. Photo restoration technology still has a lot of developments. Traditional image restoration technology mainly repairs the missing areas of the image based on mathematical formulas or thermal diffusion. This technology can only repair images with simple structures and small damaged areas and is difficult to apply in people’s daily lives. The emergence of deep learning technology has accelerated the pace of research on image restoration. This article will discuss the methods of repairing old photos based on deep neural networks. It is aimed at proposing an image restoration method based on deep neural network to enhance the effect of image restoration of old photos and provide more possibilities for restoration of old photos. This article discusses the background significance of image restoration methods, designs an image restoration model based on deep neural networks, and introduces the structure, principle, and loss function of the model. Finally, this article did a comparative experiment to compare the model in this article with other models and draw the conclusion: in the blur repair experiment, the algorithm in this paper is better than other algorithms for the peak signal-to-noise ratio and structure similarity of the repaired image; in the damage repair experiment, the value of the algorithm’ s peak signal-to-noise ratio is 32.34, and the structure similarity under different damage average levels is 0.767, which is also higher than other algorithms. Therefore, the model in this paper has the best effect on image restoration.

show abstract

Section: Image Restoration Processing Of Old Photosmentioning

confidence: 99%

An Old Photo Image Restoration Processing Based on Deep Neural Network Structure

Wang

2022

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

show abstract

“…Aiming to recover an underlying low rank matrix from its degraded observation, low rank matrix approximation can robustly and efficiently handle high-dimensional data with high noise or severe corruption, due to the fact that many types of data (raw or after some nonlinear transforms) reside near single or multiple subspaces [34]. Singular value decomposition is often an effective approach to solve low rank model using special thresholding operations on the singular values of observation matrix [24,31,[35][36][37].…”

Section: Gradual Reweighted Regularizationmentioning

confidence: 99%

Image Restoration Based on Gradual Reweighted Regularization and Low Rank prior

Wang

2020

Mathematical Problems in Engineering

View full text Add to dashboard Cite

Digital restoration of image with missing data is a basic need for visual communication and industrial applications. In this paper, making full use of priors of low rank and nonlocal self-similarity a gradual reweighted regularization is proposed for matrix completion and image restoration. Sparsity-promoting regularization produces much sparser representation of grouped nonlocal similar blocks of image by solving a nonconvex minimization problem. Moreover, an alternation direction method of multipliers algorithm is developed to speed up iterative solving of the above problem. Image block classification further enhances the adaptivity of the proposed method. Experiments on simulated matrix and natural image show that the proposed method obtains better image restoration results, where most lost information is reorcovered and few artifacts are produced.

show abstract

“…The WT is also well applied to noise reduction which helps to get clean images such as the wavelet combines morphology or partial differential equations to achieve denoising [20]. In addition, the wavelet algorithm selects threshold and threshold function to improve performance on restoration [21]. The Retinex algorithm is built on the theory of colour constancy based on the visual representation of human [22, 23].…”

Section: Introductionmentioning

confidence: 99%

Restoration algorithm for noisy complex illumination

Liu

Gao

Kong

et al. 2018

IET Computer Vision

View full text Add to dashboard Cite

Although promising results have been achieved in the restoration of complex illumination images with the Retinex algorithm, there are still some drawbacks in the processing of Retinex. Considering the noise characteristics of complex illumination images, in this study, we propose a novel restoration algorithm for noisy complex illumination, which combines guided adaptive multi-scale Retinex (GAMSR) and improvement BayesShrink threshold filtering (IBTF) based on double-density dual-tree complex wavelet transform (DDDTCWT) domain. Extensive restoration experiments are conducted on three typical types images and the same image with different noises. On the basis of a series of evaluation indexes, we compare our method to those of state-of-the-art algorithms. The results show that (i) SSIM of the proposed IBTF is superior to traditional Bayes threshold method by 15% as the standard variance is 100. (ii) PSNR of the proposed GAMSR enhances 15% to traditional MSR. (iii) The clarity of final results for restoration speeds up three times than that of original images, and the information entropy is improved slightly too. Therefore, the proposed method can effectively enhance the details, edges and textures of the image under complex illumination and noises.

show abstract

MF-LRTC: Multi-filters guided low-rank tensor coding for image restoration

Cited by 6 publications

References 24 publications

An Old Photo Image Restoration Processing Based on Deep Neural Network Structure

An Old Photo Image Restoration Processing Based on Deep Neural Network Structure

Image Restoration Based on Gradual Reweighted Regularization and Low Rank prior

Restoration algorithm for noisy complex illumination

Contact Info

Product

Resources

About