Deep-learning denoising computational ghost imaging

Wu, Heng; Wang, Ruizhou; Zhao, Genping; Xiao, Huifang; Liang, Jian; Wang, Daodang; Tian, Xiaobo; Cheng, Lianglun; Zhang, Xianmin

doi:10.1016/j.optlaseng.2020.106183

Cited by 68 publications

(25 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A training set with a patch group was created and then the deep learning method [94,95] was used to reduce the noise. Reference [96] developed an end-to-end deep neural network (DDANet) for computational ghost image reconstruction. DDANet used a bucket signal with multiple tunable noise-level maps.…”

Section: Cnn Denoising For Specific Imagesmentioning

confidence: 99%

Methods for image denoising using convolutional neural network: a review

Ilesanmi

2021

Complex Intell. Syst.

168

View full text Add to dashboard Cite

Image denoising faces significant challenges, arising from the sources of noise. Specifically, Gaussian, impulse, salt, pepper, and speckle noise are complicated sources of noise in imaging. Convolutional neural network (CNN) has increasingly received attention in image denoising task. Several CNN methods for denoising images have been studied. These methods used different datasets for evaluation. In this paper, we offer an elaborate study on different CNN techniques used in image denoising. Different CNN methods for image denoising were categorized and analyzed. Popular datasets used for evaluating CNN image denoising methods were investigated. Several CNN image denoising papers were selected for review and analysis. Motivations and principles of CNN methods were outlined. Some state-of-the-arts CNN image denoising methods were depicted in graphical forms, while other methods were elaborately explained. We proposed a review of image denoising with CNN. Previous and recent papers on image denoising with CNN were selected. Potential challenges and directions for future research were equally fully explicated.

show abstract

Section: Cnn Denoising For Specific Imagesmentioning

confidence: 99%

Methods for image denoising using convolutional neural network: a review

Ilesanmi

2021

Complex Intell. Syst.

168

View full text Add to dashboard Cite

show abstract

“…The element of DMD has two states of ‘0’ and ‘1’, which corresponds to different deflection angles. To realize modulation of reflected light of DMD, the deflection angles are controlled by a computer according to retina-like patterns [ 21 ]. The reflected light is collimated by the projecting lens and illuminated on the surface of object.…”

Section: Theorymentioning

confidence: 99%

“…However, the total amount of data increased with the increase of compression ratio, and the algorithms time consumption reduced with the increase of the number of blocks. Heng Wu et al [ 21 ] proposed a CGI method based on deep learning denoising under the condition of sub-Nyquist sampling ratio. This method can obtain a clear object image and has practical applications in image denoising.…”

Section: Introductionmentioning

confidence: 99%

A Novel Approach of Parallel Retina-Like Computational Ghost Imaging

Cao

Zhou

Zhang

et al. 2020

Sensors

View full text Add to dashboard Cite

Computational ghost imaging (CGI), with the advantages of wide spectrum, low cost, and robustness to light scattering, has been widely used in many applications. The key issue is long time correlations for acceptable imaging quality. To overcome the issue, we propose parallel retina-like computational ghost imaging (PRGI) method to improve the performance of CGI. In the PRGI scheme, sampling and reconstruction are carried out by using the patterns which are divided into blocks from designed retina-like patterns. Then, the reconstructed image of each block is stitched into the entire image corresponding to the object. The simulations demonstrate that the proposed PRGI method can obtain a sharper image while greatly reducing the time cost than CGI based on compressive sensing (CSGI), parallel architecture (PGI), and retina-like structure (RGI), thereby improving the performance of CGI. The proposed method with reasonable structure design and variable selection may lead to improve performance for similar imaging methods and provide a novel technique for real-time imaging applications.

show abstract

“…The use of machine learning algorithms has recently gained a lot of interest due to their promising capabilities. Deep learning has been used to increase image reconstruction quality 28 and improve image quality by denoising mechanisms 29 , 30 through the use of neural networks. To speed up and enhance image reconstructions deep convolution auto-encoders have been used as a self-supervised learning approach 31 , 32 .…”

Section: Introductionmentioning

confidence: 99%

Deep learning early stopping for non-degenerate ghost imaging

Moodley

Sephton

Rodríguez-Fajardo

et al. 2021

Sci Rep

View full text Add to dashboard Cite

Quantum ghost imaging offers many advantages over classical imaging, including the ability to probe an object with one wavelength and record the image with another (non-degenerate ghost imaging), but suffers from slow image reconstruction due to sparsity and probabilistic arrival positions of photons. Here, we propose a two-step deep learning approach to establish an optimal early stopping point based on object recognition, even for sparsely filled images. In step one we enhance the reconstructed image after every measurement by a deep convolutional auto-encoder, followed by step two in which a classifier is used to recognise the image. We test this approach on a non-degenerate ghost imaging setup while varying physical parameters such as the mask type and resolution. We achieved a fivefold decrease in image acquisition time at a recognition confidence of $$75\%$$ 75 % . The significant reduction in experimental running time is an important step towards real-time ghost imaging, as well as object recognition with few photons, e.g., in the detection of light sensitive structures.

show abstract

Deep-learning denoising computational ghost imaging

Cited by 68 publications

References 31 publications

Methods for image denoising using convolutional neural network: a review

Methods for image denoising using convolutional neural network: a review

A Novel Approach of Parallel Retina-Like Computational Ghost Imaging

Deep learning early stopping for non-degenerate ghost imaging

Contact Info

Product

Resources

About