Abstract

An image interpolation model based on sparse representation is proposed. Two widely used priors including sparsity and nonlocal self-similarity are used as the regularization terms to boost the performance of the interpolation model. Meanwhile, we incorporate nonlocal linear regression into this model, since nonlocal similar patches could provide a better approximation to a given patch. Moreover, we propose a new approach to learn an adaptive sub-dictionary online instead of clustering. For each patch, similar patches are grouped to learn the adaptive sub-dictionary, generating a more sparse and accurate representation. Finally, weighted encoding is introduced to suppress tailing of fitting residuals in data fidelity. Abundant experimental results show that our proposed method achieves better performance compared to several state-of-the-art methods in terms of subjective and objective evaluations.

© 2020 Optical Society of America

Image fusion via nonlocal sparse K-SVD dictionary learning

Ying Li, Fangyi Li, Bendu Bai, and Qiang Shen
Appl. Opt. 55(7) 1814-1823 (2016)

Single image super-resolution using locally adaptive multiple linear regression

Soohwan Yu, Wonseok Kang, Seungyong Ko, and Joonki Paik
J. Opt. Soc. Am. A 32(12) 2264-2275 (2015)

Computational ghost imaging via adaptive deep dictionary learning

Xiang Zhai, Zhengdong Cheng, Zhenyu Liang, Yi Chen, Yangdi Hu, and Yuan Wei
Appl. Opt. 58(31) 8471-8478 (2019)

Sparse representations for online-learning-based hyperspectral image compression

İrem Ülkü and Behçet Uğur Töreyin
Appl. Opt. 54(29) 8625-8631 (2015)

Polarization prior to single-photon counting image denoising

Xin Tian, Wei Chen, Zhongyuan Wang, and Jiayi Ma
Opt. Express 29(14) 21664-21682 (2021)