Abstract

Infrared image denoising is an essential inverse problem that has been widely applied in many fields. However, when suppressing impulse noise, existing methods lead to blurred object details and loss of image information. Moreover, computational efficiency is another challenge for existing methods when processing infrared images with large resolution. An infrared image impulse-noise-suppression method is introduced based on tensor robust principal component analysis. Specifically, we propose a randomized tensor singular-value thresholding algorithm to solve the tensor kernel norm based on the matrix stochastic singular-value decomposition and tensor singular-value threshold. Combined with the image blocking, it can not only ensure the denoising performance but also greatly improve the algorithm’s efficiency. Finally, truncated total variation is applied to improve the smoothness of the denoised image. Experimental results indicate that the proposed algorithm outperforms the state-of-the-art methods in computational efficiency, denoising effect, and detail feature preservation.

© 2021 Optical Society of America

Graph-regularized tensor robust principal component analysis for hyperspectral image denoising

Yongming Nie, Linsen Chen, Hao Zhu, Sidan Du, Tao Yue, and Xun Cao
Appl. Opt. 56(22) 6094-6102 (2017)

General demosaicking for multispectral polarization filter arrays using total generalized variation and weighted tensor nuclear norm minimization

Kazuma Shinoda, Kota Yokoyama, and Madoka Hasegawa
Appl. Opt. 60(20) 5967-5976 (2021)

Hyperspectral image denoising using the robust low-rank tensor recovery

Chang Li, Yong Ma, Jun Huang, Xiaoguang Mei, and Jiayi Ma
J. Opt. Soc. Am. A 32(9) 1604-1612 (2015)

BM3D adaptive TV filtering-based convolutional neural network for ESPI image denoising

Huamei Xin, Zengzhao Sun, Ying Xing, and Jingjing Wang
Appl. Opt. 60(35) 10920-10927 (2021)

Polarization prior to single-photon counting image denoising

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