Noise Removal Algorithm Research Articles

A Nested UNet Based on Multi-Scale Feature Extraction for Mixed Gaussian-Impulse Removal

Eliminating mixed noise from images is a challenging task because accurately describing the attenuation of noise distribution is difficult. However, most existing algorithms for mixed noise removal solely rely on the local information of the image and neglect the global information, resulting in suboptimal denoising performance when dealing with complex mixed noise. In this paper, we propose a nested UNet based on multi-scale feature extraction (MSNUNet) for mixed noise removal. In MSNUNet, we introduce a U-shaped subnetwork called MSU-Subnet for multi-scale feature extraction. These multi-scale features contain abundant local and global features, aiding the model in estimating noise more accurately and improving its robustness. Furthermore, we introduce a multi-scale feature fusion channel attention module (MSCAM) to effectively aggregate feature information from different scales while preserving intricate image texture details. Our experimental results demonstrate that MSNUNet achieves leading performance in terms of quality metrics and the visual appearance of images.

Accurate Detection for Zirconium Sheet Surface Scratches Based on Visible Light Images.

Zirconium sheet has been widely used in various fields, e.g., chemistry and aerospace. The surface scratches on the zirconium sheets caused by complex processing environment have a negative impact on the performance, e.g., working life and fatigue fracture resistance. Therefore, it is necessary to detect the defect of zirconium sheets. However, it is difficult to detect such scratch images due to lots of scattered additive noise and complex interlaced structural texture. Hence, we propose a framework for adaptively detecting scratches on the surface images of zirconium sheets, including noise removing and texture suppressing. First, the noise removal algorithm, i.e., an optimized threshold function based on dual-tree complex wavelet transform, uses selected parameters to remove scattered and numerous noise. Second, the texture suppression algorithm, i.e., an optimized relative total variation enhancement model, employs selected parameters to suppress interlaced texture. Finally, by connecting disconnection based on two types of connection algorithms and replacing the Gaussian filter in the standard Canny edge detection algorithm with our proposed framework, we can more robustly detect the scratches. The experimental results show that the proposed framework is of higher accuracy.

An adaptive optimum weighted mean filter and bilateral filter for noise removal in cardiac MRI images

Medical diagnosis greatly benefits from the use of MRI images. There have been many MRI image enhancement algorithms proposed over the years to aid doctors in disease diagnosis. Conventional image enhancing methods, nevertheless, may also boost the noise that was already there in the captured picture, which might cause distortion, which is unfavourable for the disease diagnosis. Therefore, an appropriate technique for noise suppression is necessary. An attempt has been made to incorporate a two-step filtering based algorithm for noise removal in cardiac MRI images while maintaining the images crucial compositional elements, such as their borders and textures. A combination of adaptive optimum weighted mean filter (AOWMF) and bilateral filter (BF) has been implemented to reduce impulsive noise and Gaussian noise that are both present as mixed noise. The fundamental idea of the AOWMF approach includes the detection of noise pixels using Crow Optimization Algorithm (COA), and replacing them with an optimum weighted mean value depending on a criterion of maximization fitness function. Peak Signal to Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) have been used to assess the effectiveness of the employed noise reduction technique. The result analysis are experimented in MATLB which reveals that proposed method is able to achieve high PSNR and SSIM and further, this method has the potential to enhance more detail structures of the input cardiac MRI images than existing methods.

IR-VIO: Illumination-Robust Visual-Inertial Odometry Based on Adaptive Weighting Algorithm With Two-Layer Confidence Maximization

Illumination change, image blur, and fast motion dramatically decrease the performance of visual-inertial navigation systems (VINS). This article presents a new illumination-robust visual-inertial odometry (IR-VIO) based on adaptive weighting algorithm with two-layer confidence maximization. First, to prevent the VIO performance degradation caused by poor image quality in complex scenes and ignoring the confidence differences of feature points, we develop a novel adaptive weighting algorithm on the multisensor layer and visual feature layer to better fuse multisensor information and maximize the overall confidence of VIO. Second, to solve the problems of image feature tracking difficulty and excessive image noise in illumination-changing scenes, an image enhancement algorithm is introduced to enhance consecutive images to the same brightness level, while a block noise removal algorithm with constraint protection mechanism is proposed to dynamically remove noise points. Finally, experimental results in the public dataset and real-world environments demonstrate that IR-VIO has superior performance in terms of accuracy and robustness compared with the state-of-the-art methods.

Adaptive myriad filtering algorithms for removal of nonstationary noise in electrooculograms

Suppressing nonstationary noise present in biomedical signals is important to provide high-quality diagnoses. Nonstationary noise is difficult for removing due to its time-varying and previously unknown characteristics. The application of linear filtering to the electrooculograph (EOG) signals leads to the smoothing of diagnostically important rapid changes in a signal caused by saccadic eye movements. In this respect, for processing edges and other discontinuous transitions, nonlinear filters based on robust estimators are more appropriate. The paper introduces novel adaptive algorithms for real-time nonlinear filtering of nonstationary noise in EOG signal with a noise- and signal-dependent filter switching, which is more appropriate for processing a local vicinity of the current input signal sample. One of the algorithms is based on myriad filters and sub-filter weighted FIR (which inite Impulse Responce) myriad hybrid filters. It suggests replacing the median with a myriad operation, calculated by Newton¢s numerical technique with adaptive switching of window length and linearity parameter settings. The other algorithm adaptively switches sub-filter weighted FIR median hybrid and averaging filters with different window lengths, offering simpler calculations and high-speed performance. These algorithms do not require time for filter parameters modification and their exact tuning during real-time signal processing and a prior knowledge of the signal model and noise variance. Numerical simulations were conducted to evaluate the filtering quality based on criteria of mean-square error and signal-to-noise ratio for a model signal under different levels of Gaussian noise. The achieved results show good performance and algorithm high quality for suppression of nonstationary noise in EOG. The myriad type adaptive algorithm prevails over the median in effectiveness but requires a numerical technique for cost function minimization, however, myriad filtering real-time implementation is possible with utilization of high-speed computers. Suggested adaptive algorithms significantly improve the efficiency of nonadaptive filters.

Decomposed Dissimilarity Measure for Evaluation of Digital Image Denoising

A new approach to the evaluation of digital image denoising algorithms is presented. In the proposed method, the mean absolute error (MAE) is decomposed into three components that reflect the different cases of denoising imperfections. Moreover, aim plots are described, which are designed to be a very clear and intuitive form of presentation of the new decomposed measure. Finally, examples of the application of the decomposed MAE and the aim plots in the evaluation of impulsive noise removal algorithms are presented. The decomposed MAE measure is a hybrid of the image dissimilarity measure and detection performance measures. It provides information about sources of errors such as pixel estimation errors, unnecessary altered pixels, or undetected and uncorrected distorted pixels. It measures the impact of these factors on the overall correction performance. The decomposed MAE is suitable for the evaluation of algorithms that perform a detection of the distortion that affects only a certain fraction of the image pixels.

A methodological framework for specular return removal from photon-counting LiDAR data

The Ice, Cloud, and land Elevation Satellite-2 (ICESat-2) data may contain abundant noise photons induced by specular reflection, which make the signal photon detection and surface elevation extraction challenging, especially on flat surfaces with high reflectivity. However, no efficient algorithm for specular return noise removal has been proposed. Therefore, this study aims to propose a novel methodological framework based on statistical features to remove noise photons induced by specular reflection. First, ICESat-2 data preprocessing was performed to improve the speed and accuracy of subsequent processing. Second, the statistical features of outlier distance were calculated and utilized to filter out noise photons. Third, the specular return noise was removed according to the extreme points in the elevation distribution histogram. Finally, the manually labeled of photons and in-situ data were utilized to evaluate the performance of the proposed framework. Additionally, the proposed framework was compared with three existing signal photon extraction algorithms (ATL03, ATL13, AVEBM) for further assessment. The experimental results show that the F1-score of the proposed framework is 99.9 % and the accuracy of water level estimation is extremely high (bias = 0.041 m, and RMSE = 0.082 m). Further analysis demonstrates that our framework has the high robustness because it is not sensitive to input parameter. Meanwhile, the weak beam achieved higher water elevation accuracy than that of the strong beam. Besides, the proposed framework performs better than the existing methods in extracting water surface photons and water level. All in all, the proposed framework has been shown to effectively remove noise photons of specular returns in the raw ICESat-2 data and reduce the underestimation of the water level.

Superpixel-Based Stripe Noise Removal for Satellite Imageries

This work introduces a novel noise removal algorithm for satellite imageries based on superpixel segmentation followed by statistics-based filtering. The algorithm worked in three main steps. First, the noisy input image was divided into subregions by employing simple linear iterative clustering (SLIC)-based superpixel segmentation. Then, the statistical property of each subregion was calculated, including their standard deviations and maximum values. Last, an adaptive statistics-based stripe noise removal was performed for each subregion by constructing adaptive filter sizes according to calculated properties. The algorithm was tested using real satellite imageries taken by the LAPAN-A2 and LAPAN-A3 satellites. Its performance was then compared to three existing methods in terms of image quality and computation speed. Extensive experiments on two datasets of 3-channel images captured by the LAPAN-A2 satellite showed that the algorithm was capable of reducing the stripe pattern as measured using the peak-signal-to-noise-ratio (PSNR) metric without introducing additional artifacts, which commonly appeared on over-corrected regions. Moreover, compared to existing methods, the proposed algorithm ran 42 to 103 times faster and provided better image quality by 2.46%, measured using the structural similarity metric (SSIM). The code of this work and the datasets used for the testing are publicly available on www.github.com/dancingpixel/SPSNR.

Abstract A22: Complete mutation concordance with amplicon-based NGS between peripheral blood and bone marrow at 0.5% VAF threshold

Abstract Background: Acute myeloid leukemia (AML) is a hematological malignancy characterized by genomic heterogeneity throughout disease progression and between patients. Traditionally, both multiparameter flow cytometry (MFC) and next-generation sequencing (NGS) samples are collected via bone marrow (BM) aspiration, which is an invasive procedure with risk of infection. Peripheral blood (PB) has been demonstrated as a good substitute for BM but the concordance at different levels of NGS variant allele frequency (VAF) has not been defined. This study aims to demonstrate the relationship of NGS signals between PB and BM. Method: Genomic profiles of 60 matched samples of BM and PB were obtained retrospectively from 24 patients with de novo AML (n=17) or secondary AML (n=7). The samples were subjected to library preparation with a hematologic panel based on AmpliMark, an ultrasensitive amplicon-based NGS platform technology. The hematologic panel covers up to 45 genes common in myeloid and lymphoid neoplasms, with an established limit of detection of 0.1% VAF. Optimized bioinformatics pipeline and in-house sequencing noise removal algorithm were applied for variant selection. One of the secondary AML patients had a concurrent diagnosis of multiple myeloma; for the purposes of this analysis, the myeloma-associated variants were excluded. Results: 50 out of 60 matched BM and PB samples (83.3%) in the cohort were positive for at least one variant in either sample type. Complete or partial concordance of genetic profiles was achieved in 86.7% of the matched BM and PB samples. All samples obtained during diagnosis or relapse showed complete concordance, while all partial concordant and discordant profiles were sampled during post-treatment or remission stages. All BM samples with ≥1.5% abnormal cells detected by MFC demonstrated complete concordance between BM and PB genetic profiles. Analysis at the individual variant level (n=115) showed 78.6% of positive percent agreement (PPA) comparing PB to BM in variant detection. For variants detected at ≥0.5% and ≥0.1% VAF in the BM samples, 100% PPA and 86.4% PPA were achieved from matched PB samples, respectively. A strong correlation of VAF (r=0.98) was also demonstrated between variants detected in BM and PB. Conclusion: PB NGS is particularly sensitive for detection of myeloid specific mutations in de novo and secondary AMLs with complete concordance at VAFs ≥0.5%, when compared to BM. The high concordance and minimally invasive nature of PB sampling makes it a favorable option for diagnostic and monitoring of AML. Citation Format: Benjamin Lim, Chae Yin Cher, Jonathan Poh, Kao Chin Ngeow, Zi Yi Lim, Min-Han Tan. Complete mutation concordance with amplicon-based NGS between peripheral blood and bone marrow at 0.5% VAF threshold [abstract]. In: Proceedings of the AACR Special Conference: Acute Myeloid Leukemia and Myelodysplastic Syndrome; 2023 Jan 23-25; Austin, TX. Philadelphia (PA): AACR; Blood Cancer Discov 2023;4(3_Suppl):Abstract nr A22.

Skeleton-based noise removal algorithm for binary concrete crack image segmentation

Image processing methods for automated concrete crack detection are often challenged by binary noise. Noise removal methods decrease the false positive pixels of crack detection results, often at the cost of a reduction in true positives. This paper proposes a novel method for binary noise removal and segmentation of noisy concrete crack images. The method applies an area threshold before reducing the pixel groups in the image to a skeleton. Each skeleton is connected to its nearest neighbour before the remaining short skeletons in the image are removed using a length threshold. A morphological reconstruction follows to remove all elements in the original noisy image that do not intersect with the skeleton. Finally, pixel groups in close proximity to the endpoints of the pixel groups in the resulting image are reinstated. Testing was conducted on a dataset of noisy binary crack images; the proposed method (Skele-Marker) obtained recall, precision, intersection over union, and F1 score results of 77%, 91%, 72%, and 84%, respectively. Skele-Marker was compared to other methods found in literature and was found to outperform other methods in terms of precision, intersection over union and F1 score. The proposed method is used to make crack detection results more reliable, supporting the ever-growing demand for automated inspections of concrete structures.

Accuracy and Precision of Iodine Quantification in Subtracted Micro-Computed Tomography: Effect of Reconstruction and Noise Removal Algorithms

PurposeTo evaluate the effect of reconstruction and noise removal algorithms on the accuracy and precision of iodine concentration (CI) quantified with subtracted micro-computed tomography (micro-CT).ProceduresTwo reconstruction algorithms were evaluated: a filtered backprojection (FBP) algorithm and a simultaneous iterative reconstruction technique (SIRT) algorithm. A 3D bilateral filter (BF) was used for noise removal. A phantom study evaluated and compared the image quality, and the accuracy and precision of CI in four scenarios: filtered FBP, filtered SIRT, non-filtered FBP, and non-filtered SIRT. In vivo experiments were performed in an animal model of chemically-induced mammary cancer.ResultsLinear relationships between the measured and nominal CI values were found for all the scenarios in the phantom study (R2 > 0.95). SIRT significantly improved the accuracy and precision of CI compared to FBP, as given by their lower bias (adj. p-value = 0.0308) and repeatability coefficient (adj. p-value < 0.0001). Noise removal enabled a significant decrease in bias in filtered SIRT images only; non-significant differences were found for the repeatability coefficient. The phantom and in vivo studies showed that CI is a reproducible imaging parameter for all the scenarios (Pearson r > 0.99, p-value < 0.001). The contrast-to-noise ratio showed non-significant differences among the evaluated scenarios in the phantom study, while a significant improvement was found in the in vivo study when SIRT and BF algorithms were used.ConclusionsSIRT and BF algorithms improved the accuracy and precision of CI compared to FBP and non-filtered images, which encourages their use in subtracted micro-CT imaging.

A noise removal algorithm based on adaptive elevation difference thresholding for ICESat-2 photon-counting data

Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) has shown great potential for near-shore bathymetry. Unavoidably, the photon data acquired has a lot of noise. In order to remove noise more accurately, a photon denoising algorithm based on adaptive elevation difference thresholding (AEDTA) is proposed under the assumption that the terrain is continuously changing. The algorithm uses a Gaussian function obtained by fitting an elevation histogram to ICESat-2 data for the extraction of underwater photons. Repeatedly count the elevation difference histogram of underwater photons and adaptively determine the threshold. And the final denoised underwater signal photons are obtained. Compared with the conventional clustering method of ordering points to identify the clustering structure (OPTICS), our algorithm achieves better photon detection results for underwater topography for different densities of ICESat-2 data and has the advantage of parameter adaption without human intervention. ICESat-2 data from Ganquan Island, Dongdao Island and Oahu Island were selected for testing, and the ICESat-2 bathymetry results calculated by the algorithm were compared with in situ bathymetry data to assess the reliability and accuracy of the proposed denoising algorithm. The results show that the MAE of the bathymetry results calculated using the ICESat-2 signal photon data extracted by the method ranged from 0.54 to 1.57 m and the RMSE ranged from 0.64 to 1.74 m in different seas. The algorithm can provide high quality denoised data for bathymetric mapping around coastal areas and islands. Meanwhile, the AEDTA algorithm also has a good denoising effect on terrestrial photons, and the effect of cloud shading on the Gaussian fit analysis of ICESat-2 data cannot be ignored.

Salt and pepper noise removal method based on graph signal reconstruction

Salt and Pepper (SAP) noise is a very common type of impulse noise which frequently arises in image acquisition and transmission. Removing SAP noise is an important task in image restoration. In this paper, we model the SAP removal problem as a graph signal reconstruction problem, so that the information of noisy pixels can also participate in the estimation of their gray values. In the proposed model, the regularization term depends on the Power Spectral Density (PSD) of original signals, which makes the reconstructed signal follows the frequency content of the potential original signal well. Further, an algorithm for estimating the local PSD of an image block is designed and corresponding theoretical analysis is given. Finally, a block SAP noise removal algorithm is proposed. The computational complexity is also discussed. Experimental results on ten commonly used test images demonstrate that the proposed algorithm outperforms the state-of-the-art methods.

Fractional-order Diffusion based Image Denoising Model

Edge indicating operators such as gradient, mean curvature, and Gauss curvature-based image noise removal algorithms are incapable of classifying edges, ramps, and flat areas adequately. These operators are often affected by the loss of fine textures. In this paper, these problems are addressed and proposed a new coefficient of diffusion for noise removal. This new coefficient consists of two edge indicating operators, namely fractional-order difference curvature and fractional-order gradient. The fractional-order difference curvature is capable of analyzing flat surfaces, edges, ramps, and tiny textures. The fractional-order gradient can able to distinguish texture regions. The selection of the order is more flexible for the fractional order gradient and fractional-order difference curvature. This will result in effective image denoising. Since the discrete Fourier transform is simple to numerically implement, it is taken into consideration for the implementation of fractional-order gradient. The proposed method can give results that are visually appealing and improved quantitative outputs in terms of the Figure of Merit (FoM), Mean Structural Similarity (MSSIM), and Peak Signal to Noise Ratio (PSNR), according to comparative analysis.

AWGN Removal Using Modified Steering Kernel and Image Matching

Image noise occurs during acquisition and transmission and adversely affects processes, such as image segmentation and object recognition and classification. Various techniques are being studied for noise removal, and with the recent development of hardware and image processing algorithms, noise removal techniques that combine non-local techniques are attracting attention. However, one disadvantage of this method is that blurring occurs in the edges and boundary line of the resulting image. In this study, we proposed a modified local steering kernel based on image matching to improve these shortcomings. The proposed technique uses image matching to differentiate the weight obtained by the steering kernel according to the local characteristics of the image and calculates the weight of the filter based on the similarity between the center window and the matching window. The resulting images were quantitatively evaluation and enlargement of images were used and compared with the existing noise removal algorithms. The proposed algorithm showed clearer contrast in the enlarged images and better results than the conventional image restoration techniques in the quantitative evaluation using peak signal-to-noise ratio and structural similarity index.

A review of recent developments in low-frequency ultra-wideband microwave radiometry for studies of the cryosphere

Over the past decade, a series of airborne experiments in the Arctic and Antarctica explored microwave emission from sea ice and ice sheets at frequencies from 0.5 to 2 GHz. The experiments were motivated by the fact that lower frequencies penetrate deeper into a frozen surface, thus offering the possibility to measure physical temperatures at great depths in ice sheets and, subsequently, other unique geophysical observables including sea ice salinity. These experiments were made feasible by recent engineering advances in electronics, antenna design, and noise removal algorithms when operating outside of protected bands in the electromagnetic spectrum. These technical advances permit a new type of radiometer that not only operates at low frequency, but also obtains continuous spectral information over the band from 0.5 to 2 GHz. Spectral measurements facilitate an understanding of the physical processes controlling emission and also support the interpretation of results from single frequency instruments. This paper reviews the development of low-frequency, wide band radiometry and its application to cryosphere science over the past 10 years. The paper summarizes the engineering design of an airborne instrument and the associated algorithms to mitigate radio frequency interference. Theoretical models of emission built around the morphologic and electrical properties of cryospheric components are also described that identify the dominant physical processes contributing to emission spectra. New inversion techniques for geophysical parameter retrieval are summarized for both Arctic and Antarctic scenarios. Examples that illustrate how the measurements are used to inform on glaciological problems are presented. The paper concludes with a description of new instrument concepts that are foreseen to extend the technology into operation from space.

Plug-and-Play-Based Algorithm for Mixed Noise Removal with the Logarithm Norm Approximation Model

During imaging and transmission, images are easily affected by several factors, including sensors, camera motion, and transmission channels. In practice, images are commonly corrupted by a mixture of Gaussian and impulse noises, further complicating the denoising problem. Therefore, in this work, we propose a novel mixed noise removal model that combines a deterministic low-rankness prior and an implicit regularization scheme. In the optimization model, we apply the matrix logarithm norm approximation model to characterize the global low-rankness of the original image. We further adopt the plug-and-play (PnP) scheme to formulate an implicit regularizer by plugging an image denoiser, which is used to preserve image details. The above two building blocks are complementary to each other. The mixed noise removal algorithm is thus established. Within the framework of the PnP scheme, we address the proposed optimization model via the alternating directional method of multipliers (ADMM). Finally, we perform extensive experiments to demonstrate the effectiveness of the proposed algorithm. Correspondingly, the simulation results show that our algorithm can recover the global structure and detailed information of images well and achieves superior performance over competing methods in terms of quantitative evaluation and visual inspection.

A novel statistical golden ratio based adaptive high density impulse noise removal algorithm

A novel statistical golden ratio based adaptive high density impulse noise removal algorithm

Hyperspectral image denoising and destriping based on sparse representation, graph Laplacian regularization and stripe low-rank property

During the acquisition of a hyperspectral image (HSI), it is easily corrupted by many kinds of noises, which limits the subsequent applications. For decades, numerous HSI denoising methods have been proposed. However, these methods rarely consider the stripe noise as an independent component, thus cannot effectively remove the stripe noise. In this paper, we propose a mixed noise removal algorithm to destripe an HSI by taking advantage of the low-rank property of stripe noise. In the meantime, sparse representation and graph Laplacian regularization are utilized to remove Gaussian and sparse noise. Roughly speaking, the sparse representation helps achieve the approximation of the original image. A graph Laplacian regularization term can ensure the non-local spatial similarity of an HSI. Separate constraints on the sparse coefficient matrix and stripe noise components can help remove different types of noises. Experimental results on both simulated and real data demonstrate the effectiveness of the proposed method for HSI restoration.

Tropical algebra for noise removal and optimal control

Algorithms for noise removal are either complex or ineffective, and the optimal control with inequality constrains makes the algorithm even more complex. Now the condition is changed completely, and the tropical algebra is an extremely simple tool for this purpose. The tropical algebra-based filter has obvious advantages over traditional ones, and an inequality constrain can be converted to a tropical polynomial, making the tropical algebra much attractive in engineering applications. In this paper, idempotents on multiplicative semigroups of tropical matrices are studied. First, the concepts of tropical algebra and the semigroup of tropical matrices under multiplication are introduced. Second, the structure of idempotents on the semigroup of [Formula: see text] tropical matrices under multiplication is given. Finally, as examples, the tropical addition is used as a filter for noise removal, and a simple optimization is given where the constraint is given by the tropical algebra. The paper has opened the path for a new way to noise removal and optimization.