On designing efficient superresolution algorithms by regression models

Kuo, Tse Ming; Tai, Shen-Chuan

doi:10.1117/1.jei.22.3.033002

Cited by 3 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Surprisingly, the listed numerical results seem to be a violation of the claim in Ref. 7, since the simple bilinear method outperforms others. However, we notice that the SR results on images down sampled by bicubic are always sharper than the results by bilinear method, which, we believe, causes the degradation of objective performance of the bicubic method since overestimation may occur in some highly textured regions.…”

Section: Discussion On Down-sampling Methodsmentioning

confidence: 50%

“…7, the down-sampling method also affects the SR algorithms, because in the synthesis experiments, the LR input images are usually obtained by down sampling the original image. Therefore, if a proper downsampling method is applied, the LR image will have fewer artifacts, which lead to a better reconstructed HR image.…”

Section: Discussion On Down-sampling Methodsmentioning

confidence: 99%

“…Because these linear kernels are usually derived under the assumption that natural images are either spatially smooth or band-limited, the interpolationbased SR methods tend to produce smooth HR images with several visual artifacts, such as ringing, aliasing, blocking, and blurry. 2,4 In recent years, some sophisticated methods, [2][3][4][5][6][7][8] in which the kernel coefficients are locally adaptive, have been proposed to reduce these unwanted artifacts. For instance, the regression model used in Refs.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Single-image superresolution based on local regression and nonlocal self-similarity

Luo

2014

J. Electron. Imaging

View full text Add to dashboard Cite

Abstract. The challenge of learning-based superresolution (SR) is to predict the relationships between low-resolution (LR) patches and their corresponding high-resolution (HR) patches. By learning such relationships from external training images, the existing learning-based SR approaches are often affected by the relevance between the training data and the LR input image. Therefore, we propose a single-image SR method that learns the LR-HR relations from the given LR image itself instead of any external images. Both the local regression model and nonlocal patch redundancy are exploited in the proposed method. The local regression model is employed to derive the mapping functions between self-LR-HR example patches, and the nonlocal self-similarity gives rise to a high-order derivative estimation of the derived mapping function. Moreover, to fully exploit the multiscale similarities inside the LR input image, we accumulate the previous reconstruction results and their corresponding LR versions as additional example patches for the subsequent estimation process, and adopt a gradual magnification scheme to achieve the desired zooming size step by step. Extensive experiments on benchmark images have validated the effectiveness of the proposed method. Compared to other state-of-the-art SR approaches, the proposed method provides photorealistic HR images with sharp edges. © The Authors.Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

show abstract

Section: Discussion On Down-sampling Methodsmentioning

confidence: 50%

Section: Discussion On Down-sampling Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Single-image superresolution based on local regression and nonlocal self-similarity

Luo

2014

J. Electron. Imaging

View full text Add to dashboard Cite

show abstract

“…From the view of calculation, each orientation pattern can be expressed in the formula form as shown in Eqs. (11) to (18) and, respectively, correspond to 0, 90, 22.5, 45, 67.5, 112.5, 135, and 157.5 deg patterns. That is, for a 5 × 5 block centered at ði; jÞ, it will satisfy the corresponding formula if it is matched by a certain orientation pattern.…”

Section: Classification Of Directional Edgesmentioning

confidence: 98%

“…Rather than the conventional regressionbased image interpolation algorithms, where the objective functions were optimized by ordinary least squares, Liu et al 10 proposed a novel image interpolation algorithm based on regularized local linear regression, which performed well in image edge structure preservation. An efficient SR algorithm based on multiple linear regression was proposed by Kuo and Tai,11 where the concept of self-similarity of a pair of LR and HR images is applied to establish the multiple linear regression model and to properly estimate the details of the HR images. In the waveletbased algorithms, the LR image is usually treated as the low-frequency wavelet subband of the wavelet-transformed HR image.…”

Section: Introductionmentioning

confidence: 99%

Artifact-free superresolution algorithm with natural texture preservation

Tai

Kuo

Chen

2014

J. Electron. Imaging

Self Cite

View full text Add to dashboard Cite

Superresolution (SR) algorithms have recently become a hot research topic. The main purpose of image upscaling is to obtain high-resolution images from low-resolution ones, and these upscaled images should look like they had been taken with a camera having a resolution the same as the upscaled images, and at least present natural textures. In general, some SR algorithms preserve clear edges but blur the textures, while others preserve detailed textures but cause some obvious artifacts along edges. The proposed SR algorithm presents the detailed textures and, meanwhile, refines the strong edges and avoids causing obvious artifacts. The goal is achieved by using orthogonal fractal as the preliminary upscaling method in conjunction with the proper postprocessing where directional enhancement is adopted. In fact, the postprocessing part in the proposed SR algorithm can effectively reduce most jagged artifacts caused by SR algorithms. In the simulation results, it is shown that the proposed SR algorithm performs well in both objective and subjective measurements. Moreover, most detailed textures are properly enhanced and most jagged artifacts caused by SR algorithms can also be effectively reduced. © 2014 SPIE and IS&T [

show abstract

An Improved Super-Resolution Reconstruction Algorithm Based on Regularization

Wang

Dong

et al. 2013

2013 International Conference on Information Science and Cloud Computing Companion

View full text Add to dashboard Cite

On designing efficient superresolution algorithms by regression models

Cited by 3 publications

References 0 publications

Single-image superresolution based on local regression and nonlocal self-similarity

Single-image superresolution based on local regression and nonlocal self-similarity

Artifact-free superresolution algorithm with natural texture preservation

An Improved Super-Resolution Reconstruction Algorithm Based on Regularization

Contact Info

Product

Resources

About