Abstract

In an effort to improve compressive sensing and spare signal reconstruction by way of the backtracking-based adaptive orthogonal matching pursuit (BAOMP), a new sparse coding algorithm called improved adaptive backtracking-based OMP (ABOMP) is proposed in this study. Many aspects have been improved compared to the original BAOMP method, including replacing the fixed threshold with an adaptive one, adding residual feedback and support set verification, and others. Because of these ameliorations, the proposed algorithm can more precisely choose the atoms. By adding the adaptive step-size mechanism, it requires much less iteration and thus executes more efficiently. Additionally, a simple but effective contrast enhancement method is also adopted to further improve the denoising results and visual effect. By combining the IABOMP algorithm with the state-of-art dictionary learning algorithm K-SVD, the proposed algorithm achieves better denoising effects for astronomical images. Numerous experimental results show that the proposed algorithm performs successfully and effectively on Gaussian and Poisson noise removal.

© 2014 Optical Society of America

Denoising infrared maritime imagery using tailored dictionaries via modified K-SVD algorithm

L. N. Smith, C. C. Olson, K. P. Judd, and J. M. Nichols
Appl. Opt. 51(17) 3941-3949 (2012)

Denoising algorithm of OCT images via sparse representation based on noise estimation and global dictionary

Xi Zhang, Zhongliang Li, Nan Nan, and Xiangzhao Wang
Opt. Express 30(4) 5788-5802 (2022)

Dictionary learning approach for image deconvolution with variance estimation

Hang Yang, Ming Zhu, Xiaotian Wu, Zhongbo Zhang, and Heyan Huang
Appl. Opt. 53(25) 5677-5684 (2014)

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)

Sparse representations for online-learning-based hyperspectral image compression

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