Noise Reduction Techniques Research Articles

Unsupervised speckle noise reduction technique for clinical ultrasound imaging.

Deep learning-based image enhancement has significant potential in the field of ultrasound image processing, as it can accurately model complicated nonlinear artifacts and noise, such as ultrasonic speckle patterns. However, training deep learning networks to acquire reference images that are clean and free of noise presents significant challenges. This study introduces an unsupervised deep learning framework, termed speckle-to-speckle (S2S), designed for speckle and noise suppression. This framework can complete its training without the need for clean (speckle-free) reference images. The proposed network leverages statistical reasoning for the mutual training of two in vivo images, each with distinct speckle patterns and noise. It then infers speckle- and noise-free images without needing clean reference images. This approach significantly reduces the time, cost, and effort experts need to invest in annotating reference images manually. The experimental results demonstrated that the proposed approach outperformed existing techniques in terms of the signal-to-noise ratio, contrast-to-noise ratio, structural similarity index, edge preservation index, and processing time (up to 86 times faster). It also performed excellently on images obtained from ultrasound scanners other than the ones used in this work. S2S demonstrates the potential of employing an unsupervised learning-based technique in medical imaging applications, where acquiring a ground truth reference is challenging.

Smart Recognition and Authentication based on Haar Cascade Classifier and Visual Cryptography

The algorithm with an E-verification method is proposed that combines automated facial recognition technology with digital media and later it combines aspects of secret data exchange and visual cryptography. A face detection technique based on Internet of Things (IoT) selects features using the Principal Component Analysis (PCA) algorithm and the Haar cascade classifier. The selected face serving as a cover follows two level Discrete Haar Wavelet Transform. The shares generated from digital signature and fingerprint are diffused into the converted coefficients. Furthermore, the imperceptibility of the additional noise is increased by a bit-level noise reduction technique. Authenticity is confirmed by regenerating a message digest at the receiving end, and the extraction process operates in complete blindness. The approach is suitable for smart card design and may be used as an automatic recognition system in a real-world setting. Performance comparisons show a notable improvement over other approaches that are comparable. Additionally, the technique is effective against some related attacks.

Computing‐efficient video analytics for nighttime traffic sensing

AbstractThe training workflow of neural networks can be quite complex, potentially time‐consuming, and require specific hardware to accomplish operation needs. This study presents a novel analytical video‐based approach for vehicle tracking and vehicle volume estimation at nighttime using a monocular traffic surveillance camera installed over the road. To build this approach, we employ computer vision‐based algorithms to detect vehicle objects, perform vehicle tracking, and vehicle counting in a predefined detection zone. To address low‐illumination conditions, we adapt and employ image noise reduction techniques, image binary conversion, image projective transformation, and a set of heuristic reasoning rules to extract the headlights of each vehicle, pair them belonging to the same vehicle, and track moving candidate vehicle objects continuously across a sequence of video frames. The robustness of the proposed method was tested in various scenarios and environmental conditions using a publicly available vehicle dataset as well as own labeled video data.

Securing Biomedical Audio Data in IoT Healthcare Systems: An Evaluation of Encryption Methods for Enhanced Privacy

Communication technology have advanced quickly since the COVID-19 epidemic started, providing consumers with additional benefits and conveniences. Concerns over the privacy and confidentiality of this data have grown in importance as initiatives that promote the use of audio and video to enhance interpersonal interactions become more common. In the context of the Internet of Things (IoT), audio communications security is essential in the biomedical domain. Sensitive medical data may be compromised in these connections, which include exchanges between patients and doctors and broadcasts of vital signs. To protect patient privacy and reduce cybersecurity threats, strong security measures such as data encryption must be put in place. Our study attempts to address these issues in this environment. Comparative examination of the Chacha20, Salsa20, and Camellia encryption algorithms enabled us to ascertain that Chacha20 performs exceptionally well when it comes to audio file decryption and encryption speed. The results of our trials attest to this encryption method's astounding effectiveness and efficacy. We have also used the noise reduction technique, which is frequently used in audio security to enhance the quality of recordings and make it easier to identify significant information in audio signals. Then, Fourier transform technique, which is also used to analyze audio files and can be used to identify changes, extract hidden information, and authenticate audio files. By doing this, the audio files security and integrity are strengthened.

Exploring the Impact of Noise and Image Quality on Deep Learning Performance in DXA Images.

Segmentation of the femur in Dual-Energy X-ray (DXA) images poses challenges due to reduced contrast, noise, bone shape variations, and inconsistent X-ray beam penetration. In this study, we investigate the relationship between noise and certain deep learning (DL) techniques for semantic segmentation of the femur to enhance segmentation and bone mineral density (BMD) accuracy by incorporating noise reduction methods into DL models. Convolutional neural network (CNN)-based models were employed to segment femurs in DXA images and evaluate the effects of noise reduction filters on segmentation accuracy and their effect on BMD calculation. Various noise reduction techniques were integrated into DL-based models to enhance image quality before training. We assessed the performance of the fully convolutional neural network (FCNN) in comparison to noise reduction algorithms and manual segmentation methods. Our study demonstrated that the FCNN outperformed noise reduction algorithms in enhancing segmentation accuracy and enabling precise calculation of BMD. The FCNN-based segmentation approach achieved a segmentation accuracy of 98.84% and a correlation coefficient of 0.9928 for BMD measurements, indicating its effectiveness in the clinical diagnosis of osteoporosis. In conclusion, integrating noise reduction techniques into DL-based models significantly improves femur segmentation accuracy in DXA images. The FCNN model, in particular, shows promising results in enhancing BMD calculation and clinical diagnosis of osteoporosis. These findings highlight the potential of DL techniques in addressing segmentation challenges and improving diagnostic accuracy in medical imaging.

Noise reduction of stochastic density functional theory for metals.

Density Functional Theory (DFT) has become a cornerstone in the modeling of metals. However, accurately simulating metals, particularly under extreme conditions, presents two significant challenges. First, simulating complex metallic systems at low electron temperatures is difficult due to their highly delocalized density matrix. Second, modeling metallic warm-dense materials at very high electron temperatures is challenging because it requires the computation of a large number of partially occupied orbitals. This study demonstrates that both challenges can be effectively addressed using the latest advances in linear-scaling stochastic DFT methodologies. Despite the inherent introduction of noise into all computed properties by stochastic DFT, this research evaluates the efficacy of various noise reduction techniques under different thermal conditions. Our observations indicate that the effectiveness of noise reduction strategies varies significantly with the electron temperature. Furthermore, we provide evidence that the computational cost of stochastic DFT methods scales linearly with system size for metal systems, regardless of the electron temperature regime.

A Study on Handwritten Text Recognition Classification using Diverse Deep Learning Techniques and Computation of CTC Loss

Handwriting recognition encompasses the conversion of hand-written text images into digital text, where input images yield predicted textual output. Optical Character Recognition (OCR) technology has conventionally fulfilled this role. With surging mobile phone usage, leveraging text detection via mobile cameras gains significance in fields like medical script processing and exam evaluation. To enhance image quality, noise reduction techniques like binarization and thresholding are applied. Image processing entails letter segmentation and extraction. In this paper, we propose a neural network classifier model amalgamating Convolution Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks. Leveraging an existing feature dataset, the model is trained. Input images are evaluated through the neural network, yielding recognized words in a text document format. Ultimately, the predicted output is derived via Connectionist Temporal Classification (CTC)- based loss computation. Keywords: Handwritten Recognition, Deep Learning Techniques, Optical Character Recognition.

Sub-photon accuracy noise reduction of a single shot coherent diffraction pattern with an atomic model trained autoencoder

Single-shot imaging with femtosecond X-ray lasers is a powerful measurement technique that can achieve both high spatial and temporal resolution. However, its accuracy has been severely limited by the difficulty of applying conventional noise-reduction processing. This study uses deep learning to validate noise reduction techniques, with autoencoders serving as the learning model. Focusing on the diffraction patterns of nanoparticles, we simulated a large dataset treating the nanoparticles as composed of many independent atoms. Three neural network architectures are investigated: neural network, convolutional neural network and U-net, with U-net showing superior performance in noise reduction and subphoton reproduction. We also extended our models to apply to diffraction patterns of particle shapes different from those in the simulated data. We then applied the U-net model to a coherent diffractive imaging study, wherein a nanoparticle in a microfluidic device is exposed to a single X-ray free-electron laser pulse. After noise reduction, the reconstructed nanoparticle image improved significantly even though the nanoparticle shape was different from the training data, highlighting the importance of transfer learning.

Advancing Surgical Imaging with cGAN for Effective Defogging

Image-based defogging technology can significantly enhance intraoperative image quality and shows great promise in various medical fields. A new image removal algorithm based on conditional generative adversarial networks (cGAN) has been developed. This algorithm employs the Tiramisu model instead of the conventional U-Net, thereby improving its computational accuracy. Additionally, the quality of the resulting images is enhanced by incorporating more textual data. A novel visual perception method is proposed, utilizing a contrast-based approach to improve the similarity between images with the same semantic content. Experiments demonstrate that this method not only excels at fog removal but also better preserves the key visual features of the images. Compared to existing image defogging technologies, this method offers superior qualitative analysis capabilities. This advancement can aid doctors in better visualizing intraoperative images. The effectiveness and robustness of the proposed method are validated through comparative analysis with several existing image noise reduction techniques.

Impact of white noise in artificial neural networks trained for classification: Performance and noise mitigation strategies.

In recent years, the hardware implementation of neural networks, leveraging physical coupling and analog neurons has substantially increased in relevance. Such nonlinear and complex physical networks provide significant advantages in speed and energy efficiency, but are potentially susceptible to internal noise when compared to digital emulations of such networks. In this work, we consider how additive and multiplicative Gaussian white noise on the neuronal level can affect the accuracy of the network when applied for specific tasks and including a softmax function in the readout layer. We adapt several noise reduction techniques to the essential setting of classification tasks, which represent a large fraction of neural network computing. We find that these adjusted concepts are highly effective in mitigating the detrimental impact of noise.

Research on noise reduction and data mining techniques for pavement dynamic response signals

Purpose This study aims to ensure reliable analysis of dynamic responses in asphalt pavement structures. It investigates noise reduction and data mining techniques for pavement dynamic response signals. Design/methodology/approach The paper conducts time-frequency analysis on signals of pavement dynamic response initially. It also uses two common noise reduction methods, namely, low-pass filtering and wavelet decomposition reconstruction, to evaluate their effectiveness in reducing noise in these signals. Furthermore, as these signals are generated in response to vehicle loading, they contain a substantial amount of data and are prone to environmental interference, potentially resulting in outliers. Hence, it becomes crucial to extract dynamic strain response features (e.g. peaks and peak intervals) in real-time and efficiently. Findings The study introduces an improved density-based spatial clustering of applications with Noise (DBSCAN) algorithm for identifying outliers in denoised data. The results demonstrate that low-pass filtering is highly effective in reducing noise in pavement dynamic response signals within specified frequency ranges. The improved DBSCAN algorithm effectively identifies outliers in these signals through testing. Furthermore, the peak detection process, using the enhanced findpeaks function, consistently achieves excellent performance in identifying peak values, even when complex multi-axle heavy-duty truck strain signals are present. Originality/value The authors identified a suitable frequency domain range for low-pass filtering in asphalt road dynamic response signals, revealing minimal amplitude loss and effective strain information reflection between road layers. Furthermore, the authors introduced the DBSCAN-based anomaly data detection method and enhancements to the Matlab findpeaks function, enabling the detection of anomalies in road sensor data and automated peak identification.

EMD-based noise reduction study of steel cored conveyor belt containing slag signal

This paper proposes a noise reduction technique that combines empirical mode decomposition (EMD) and wavelet thresholding to eliminate the noise interference caused by metal slag in the steel cord conveyor belt damage signal. Initially, the original signal undergoes decomposition by EMD into high- and low-frequency intrinsic mode function (IMF) components based on metal slag signal feature analysis. The wavelet threshold de-noising method is then applied to filter out the high-frequency noise while preserving the low-frequency IMF components. The IMF components containing slag noise are discarded, and the filtered signal and the low-frequency IMF components are used for signal reconstruction to obtain an optimal noise reduction effect. The study demonstrates that this method is more effective in suppressing metal slag noise, environmental noise, and other noise interferences than traditional EMD decomposition noise reduction and wavelet threshold noise reduction methods. This research provides a valuable reference for noise reduction in steel cord conveyor belt detection data.

Deep learning with adaptive convolutions for classification of retinal diseases via optical coherence tomography

Optical coherence tomography (OCT) uses interferometry to capture high-resolution cross-sectional images of the retina to diagnose retinal diseases. Convolutional neural networks (CNNs) have become essential for developing efficient computer-aided diagnostic algorithms, but noisy images can hinder their performance. This study introduces an innovative image preprocessing strategy that involves a new method of representing images to reduce image noise and a new adaptive convolution layer. The adaptive convolution layer aims to replace traditional convolution layers for OCT image classification by relying on local’Feature Content’. The proposed image representation is based on Zeckendorf's theorem, which states that every positive integer may be split into a unique sum of distinct, non-adjacent Fibonacci numbers. The proposed approach enables the generation of two separate images, known as the ‘base’ and ‘fine,’ where the ‘base’ image is the denoised image. We assessed our methodology by evaluating against ten filters, comprising a Low-pass filter, Gaussian filter, Wiener filter, Wavelet filter, Guided filter, Lee filter, Frost filter, Kuan filter, Detail Preserving Anisotropic Diffusion (DPAD) filter, and Non-Local Means (NLM) filter. In our study, we found that the proposed filter produced the most favorable results in five of the six no-reference parameters (Blur Percent (BP), Blind/Referenceless Image Spatial Quality Evaluator (BRISQUE), Naturalness Image Quality Evaluator (NIQE), Wavelet Variance (WAVV), Wavelet Variance Radial (WAVR)) used to assess the effectiveness of the proposed image enhancement technique. The OCT dataset utilized in the study was compiled by the University of California San Diego. The proposed adaptive convolution layer and its accompanying activation function were tested using seven OCT image classification CNN architectures. The test architectures comprise OctNET, NT-CNN, AOCT-NET, M-CNN, LightOCT, RetiNet, and DeepOCT. Experiments were conducted to assess the impact of the new preprocessing algorithm and the placement of the adaptive convolution layer as a substitute for the conventional convolution layer. Implementing the proposed approaches resulted in accuracy improvements ranging from 0.44% to 2.44% across architectures. Our findings highlight the efficacy of the proposed indirect noise reduction technique and a texture-sensitive adaptive convolution layer.

A framework for identification of brain tumors from MR images using progressive segmentation

PurposeThis study addresses the critical health issue of brain tumors, focusing on enhancing the accuracy of tumor segmentation from Magnetic Resonance Imaging (MRI) images. The primary research question investigates the effectiveness of a novel Hybrid Watershed–Clustering framework and its underlying Progressive Segmentation of the MR Images using the Radius and Intensity Measure (PS-RIM) algorithm. The aim is to improve the detection and segmentation of brain tumors within MR images, surpassing the efficacy of current methodologies.MethodsThe methodology involves a three-stage process. In the preprocessing stage, noise reduction and intensity normalization techniques are applied to clarify the images. The next stage is region-based segmentation, which includes morphological processing, edge detection, and thresholding to delineate tumor areas accurately. The final post-processing stage enhances segmentation accuracy and reduces false positives by integrating clustering machine learning techniques, specifically the K-Means cluster algorithm, to refine tumor identification.ResultsThe framework's comprehensive evaluation across various MR images shows a significant improvement in accuracy over existing segmentation methods. The PS-RIM algorithm within the framework effectively captures the diverse presentations of tumor appearances in MR images. The research recorded an impressive accuracy rate of 98.11% in tumor detection, demonstrating enhanced identification and segmentation quality.ConclusionsThe study concludes that the proposed Hybrid Watershed–Clustering framework, powered by the PS-RIM algorithm, markedly improves the detection and differentiation of brain tumors in MR images. It exhibits exceptional accuracy, resilience, and computational efficiency. These findings hold substantial potential for advancing computer vision and image analysis in medical diagnostics, which could improve patient outcomes in managing brain tumors.Graphical abstract

Projection domain decomposition denoising algorithm based on low rank and similarity-based regularization.

Projection Domain Decomposition (PDD) is a dual energy reconstruction method which implements the decomposition process before image reconstruction. The advantage of PDD is that it can alleviate beam hardening artifacts and metal artifacts effectively as energy spectra estimation is considered in PDD. However, noise amplification occurs during the decomposition process, which significantly impacts the accuracy of effective atomic number and electron density. Therefore, effective noise reduction techniques are required in PDD. This study aims to develop a new algorithm capable of minimizing noise while simultaneously preserving edges and fine details. In this study, a denoising algorithm based on low rank and similarity-based regularization (LRSBR) is presented. This algorithm incorporates the low-rank characteristic of tensors into similarity-based regularization (SBR) framework. This method effectively addresses the issue of instability in edge pixels within the SBR algorithm and enhances the structural consistency of dual-energy images. A series of simulation and practical experiments were conducted on a dual-layer dual-energy CT system. Experiments demonstrate that the proposed method outperforms exiting noise removal methods in Peak Signal-to-noise Ratio (PSNR), Root Mean Square Error (RMSE), and Structural Similarity (SSIM). Meanwhile, there has been a notable enhancement in the visual quality of CT images. The proposed algorithm has a significantly improved noise reduction compared to other competing approach in dual-energy CT. Meanwhile, the LRSBR method exhibits outstanding performance in preserving edges and fine structures, making it practical for PDD applications.

Integrated Sensor Fusion and Multi-Modal Hierarchical Neural Network for Activity Recognition in Lower Limb Prosthetics

Activity recognition plays pivotal role in enhancing functionality for prosthetic devices, ensuring seamless integration with users' movements. However, the complexity arises from the diverse data sources, including acceleration, angular velocity, joint angles, orientation, electromyography (EMG), and marker data, necessitating a robust approach to overcome challenges in information integration. The primary challenge lies in the effective utilization of multiple sensor modalities, each with unique characteristics and potential noise sources. The proposed solution addresses this by employing advanced sensor fusion techniques, such as Kalman filtering, during data collection. Synchronization and resampling ensure temporal consistency, while noise reduction techniques, such as low-pass filters, mitigate signal distortions. To further refine the process, a hybrid optimization-based feature selection Adaptive Step Size in Marine Predators Algorithm (ASSMPA) is introduced, focusing on marker data features. ASSMPA synergizes Marine Predators Algorithm (MPA) and Pathfinder Algorithm (PFA) for optimal feature selection in marine predator pathfinding tasks. The Feature Fusion step integrates attention mechanisms to dynamically weigh the significance of different sensor modalities during the fusion process. This strategic fusion enhances the overall performance of the Multi-Modal Hierarchical Neural Network (MMHNN). The proposed model is implemented using Python.

Numerical solution and experimental investigation for hydro-acoustic analysis and noise reduction assessment of ship ducted propeller

Ship noise emission has many environmental impacts. Recognizing noise-generating source and investigating noise reduction techniques is the first step in elimination or reduction of these harmful effects. Studies show that ship propeller is one of the main sources of ship noise emission. Ducted propeller is one of the well-known solutions for noise reduction. In this paper, results of numerical solution for governing mathematical equations of noise emission in far field is presented and sound pressure level is calculated using Ffowcs Williams and Hawkings (FWH) equations. Numerical solution is validated by valid research resources. Propeller noise emission is also calculated by experimental measurements and filtering ambient noise frequencies. Finally, a comparison has been made between numerical solution and experimental results and the effect of duct on noise reduction is calculated. The measured SPL emitted from the propeller in CFD approach (solving the FWH) shows 28% reduction for ducted propeller in comparison with the propeller without the duct. In experimental approach (after removing the ambient noise) shows 37% reduction for ducted propeller in comparison with the propeller without the duct.

Integrated Sensor Fusion and Multi-Modal Hierarchical Neural Network for Activity Recognition in Lower Limb Prosthetics

Activity recognition plays pivotal role in enhancing functionality for prosthetic devices, ensuring seamless integration with users' movements. However, the complexity arises from the diverse data sources, including acceleration, angular velocity, joint angles, orientation, electromyography (EMG), and marker data, necessitating a robust approach to overcome challenges in information integration. The primary challenge lies in the effective utilization of multiple sensor modalities, each with unique characteristics and potential noise sources. The proposed solution addresses this by employing advanced sensor fusion techniques, such as Kalman filtering, during data collection. Synchronization and resampling ensure temporal consistency, while noise reduction techniques, such as low-pass filters, mitigate signal distortions. To further refine the process, a hybrid optimization-based feature selection Adaptive Step Size in Marine Predators Algorithm (ASSMPA) is introduced, focusing on marker data features. ASSMPA synergizes Marine Predators Algorithm (MPA) and Pathfinder Algorithm (PFA) for optimal feature selection in marine predator pathfinding tasks. The Feature Fusion step integrates attention mechanisms to dynamically weigh the significance of different sensor modalities during the fusion process. This strategic fusion enhances the overall performance of the Multi-Modal Hierarchical Neural Network (MMHNN). The proposed model is implemented using Python

Improved tactile speech robustness to background noise with a dual-path recurrent neural network noise-reduction method

Many people with hearing loss struggle to understand speech in noisy environments, making noise robustness critical for hearing-assistive devices. Recently developed haptic hearing aids, which convert audio to vibration, can improve speech-in-noise performance for cochlear implant (CI) users and assist those unable to access hearing-assistive devices. They are typically body-worn rather than head-mounted, allowing additional space for batteries and microprocessors, and so can deploy more sophisticated noise-reduction techniques. The current study assessed whether a real-time-feasible dual-path recurrent neural network (DPRNN) can improve tactile speech-in-noise performance. Audio was converted to vibration on the wrist using a vocoder method, either with or without noise reduction. Performance was tested for speech in a multi-talker noise (recorded at a party) with a 2.5-dB signal-to-noise ratio. An objective assessment showed the DPRNN improved the scale-invariant signal-to-distortion ratio by 8.6 dB and substantially outperformed traditional noise-reduction (log-MMSE). A behavioural assessment in 16 participants showed the DPRNN improved tactile-only sentence identification in noise by 8.2%. This suggests that advanced techniques like the DPRNN could substantially improve outcomes with haptic hearing aids. Low-cost haptic devices could soon be an important supplement to hearing-assistive devices such as CIs or offer an alternative for people who cannot access CI technology.

Stock prediction and analysis based on machine learning algorithms

The stock market has consistently remained a focal point of substantial concern for investors. Nevertheless, due to the intricate, tumultuous, and often noisy nature of the stock market, forecasting stock trends presents a formidable obstacle. To augment the accuracy of stock trend predictions, the author adopts a combination of the Long Short-Term Memory (LSTM) neural network and a noise reduction technique known as Ensemble Empirical Mode Decomposition (EEMD). This composite model is employed to develop predictions for the daily stock price increases, aiming to provide more precise insights into market behavior. The framework is capable of generating the daily stock price change trend curve based on the training outcomes. EEMD, standardization, and other data preprocessing methods can effectively reduce the noise of the stock market. In this paper, three U.S. stocks from 2010 to 2023 are chosen as the research subjects. After the training is completed, the prediction curve generated by the model closely aligns with the actual curve. Furthermore, three commonly used evaluation metrics were utilized to assess the models performance. Based on all those experimental outcomes, this model adeptly forecasts the stocks trend.