Fractional Variation Network for THz Spectrum Denoising without Clean Data

Jiao, Qingliang; Xu, Jing; Liu, Ming; Zhao, Fengfeng; Dong, Liquan; Hui, Mei; Kong, Lingqin; Zhao, Yuejin

doi:10.3390/fractalfract6050246

Cited by 4 publications

(3 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fractional calculus, with its three-century history, was long confined to the realm of pure mathematical analysis and largely ignored by engineers [1][2][3][4][5]. It was not until Mandelbrot introduced fractal theory, linking Riemann-Liouville fractional calculus to Brownian motion in fractal media [6][7][8], that the field captured the attention of engineering technologists.…”

Section: Introductionmentioning

confidence: 99%

Research on Application of Fractional Calculus Operator in Image Underlying Processing

Huang,

Qin,

Chen

et al. 2024

Fractal Fract

View full text Add to dashboard Cite

Fractional calculus extends traditional, integer-based calculus to include non-integer orders, offering a powerful tool for a range of engineering applications, including image processing. This work delves into the utility of fractional calculus in two crucial aspects of image processing: image enhancement and denoising. We explore the foundational theories of fractional calculus together with its amplitude–frequency characteristics. Our focus is on the effectiveness of fractional differential operators in enhancing image features and reducing noise. Experimental results reveal that fractional calculus offers unique benefits for image enhancement and denoising. Specifically, fractional-order differential operators outperform their integer-order counterparts in accentuating details such as weak edges and strong textures in images. Moreover, fractional integral operators excel in denoising images, not only improving the signal-to-noise ratio but also better preserving essential features such as edges and textures when compared to traditional denoising techniques. Our empirical results affirm the effectiveness of the fractional-order calculus-based image-processing approach in yielding optimal results for low-level image processing.

show abstract

Section: Introductionmentioning

confidence: 99%

Research on Application of Fractional Calculus Operator in Image Underlying Processing

Huang,

Qin,

Chen

et al. 2024

Fractal Fract

View full text Add to dashboard Cite

show abstract

“…The intrinsically high resolution of infrared linear array images complicates the task of compiling a comprehensive dataset, presenting formidable challenges. The self-supervised denoising method pioneered by Jaakko Lehtinen et al [22], along with the approach by Qingliang Jiao et al [23] designed to minimize reliance on pristine training data are particularly significant. While both methodologies cater to application requirements, the neural network model's consistent input size necessitates size preprocessing for the THz spectrum, amplifying the complexity.…”

Section: Introductionmentioning

confidence: 99%

Overcoming Periodic Stripe Noise in Infrared Linear Array Images: The Fourier-Assisted Correlative Denoising Method

Chen,

2023

Sensors

View full text Add to dashboard Cite

Infrared linear array detectors frequently experience vertical, low-frequency, and periodic stripe noise during imaging, stemming from electro-mechanical interference. Unlike conventional periodic disturbances, this interference showcases long periodicities and is uniquely columnar in orientation. Its presence, especially within the low-frequency domain, renders conventional filtering techniques ineffective and, at times, detrimental to image quality. Addressing this challenge, we introduce Fourier-Assisted Correlative Denoising (FACD), a correlation-centric denoising approach tailored for such unique interference patterns. This mechanism begins with the capture of a pure background image, inclusive of periodic noise, during the non-uniform correction phase of the infrared detector. Leveraging the noise’s frequency domain attributes, we extract a one-dimensional single-cycle noise signal. The infrared image is subsequently segmented into parts, and using the detected noise periodicity, the one-dimensional signals for each segment are computed. By leveraging the correlation between these signals and the benchmark one-dimensional noise pattern, we ascertain the noise profile within each segment. This profile is then employed for spatial domain denoising across the entire image frame. Empirical assessments confirm that the FACD outperforms contemporary denoising techniques by augmenting the peak signal-to-noise ratio by approximately 2.5 dB, underscoring its superior robustness. Furthermore, in light of its specificity to this noise model, FACD rapidly denoises high-resolution real infrared linear array scans, thus meeting the stringent real-time and resolution imperatives of advanced infrared linear array scanning apparatuses.

show abstract

“…Choi et al 42 adopted the WaveNet from the field of speech and audio for THz image denoising in the frequency domain for 1D temporal signals. To overcome limited training data, Jiao et al 43 proposed a Noise2Noise-based network for THz spectrum denoising using transfer learning from low-quality underwater images. However, deep learning has not been investigated in THz imaging for historical document analysis yet.…”

mentioning

confidence: 99%

Deep learning for terahertz image denoising in nondestructive historical document analysis

Dutta

Root

Ullmann

et al. 2022

Sci Rep

View full text Add to dashboard Cite

Historical documents contain essential information about the past, including places, people, or events. Many of these valuable cultural artifacts cannot be further examined due to aging or external influences, as they are too fragile to be opened or turned over, so their rich contents remain hidden. Terahertz (THz) imaging is a nondestructive 3D imaging technique that can be used to reveal the hidden contents without damaging the documents. As noise or imaging artifacts are predominantly present in reconstructed images processed by standard THz reconstruction algorithms, this work intends to improve THz image quality with deep learning. To overcome the data scarcity problem in training a supervised deep learning model, an unsupervised deep learning network (CycleGAN) is first applied to generate paired noisy THz images from clean images (clean images are generated by a handwriting generator). With such synthetic noisy-to-clean paired images, a supervised deep learning model using Pix2pixGAN is trained, which is effective to enhance real noisy THz images. After Pix2pixGAN denoising, 99% characters written on one-side of the Xuan paper can be clearly recognized, while 61% characters written on one-side of the standard paper are sufficiently recognized. The average perceptual indices of Pix2pixGAN processed images are 16.83, which is very close to the average perceptual index 16.19 of clean handwriting images. Our work has important value for THz-imaging-based nondestructive historical document analysis.

show abstract

Fractional Variation Network for THz Spectrum Denoising without Clean Data

Cited by 4 publications

References 45 publications

Research on Application of Fractional Calculus Operator in Image Underlying Processing

Research on Application of Fractional Calculus Operator in Image Underlying Processing

Overcoming Periodic Stripe Noise in Infrared Linear Array Images: The Fourier-Assisted Correlative Denoising Method

Deep learning for terahertz image denoising in nondestructive historical document analysis

Contact Info

Product

Resources

About