Noise removal from image data using recursive neurofuzzy filters

2000 ◽  
Vol 49 (2) ◽  
pp. 307-314 ◽  
Author(s):  
F. Russo
Keyword(s):  
Image Data ◽  
Noise Removal

scholarly journals Image Enhancement and Denoising in Extreme Low-Light Conditions

2019 ◽  
Vol 9 (1) ◽  
pp. 5259-5264
Keyword(s):  
Image Enhancement ◽  
Smart Phone ◽  
Image Data ◽  
Structural Similarity ◽  
Noise Removal ◽  
Low Light ◽  
Daily Experience ◽  
Quantitative Results ◽  
Loss Term

Image noise refers to the specks of false colors or artifacts that diminish the visual quality of the captured image. It has become our daily experience that with affordable smart-phone cameras we can capture high clarity photos in a brightly illuminated scene. But using the same camera in a poorly lit environment with high ISO settings results in images that are noisy with irrelevant specks of colors. Noise removal and contrast enhancement in images have been extensively studied by researchers over the past few decades. But most of these techniques fail to perform satisfactorily if the images are captured in an extremely dark environment. In recent years, computer vision researchers have started developing neural network-based algorithms to perform automated de-noising of images captured in a low-light environment. Although these methods are reasonably successful in providing the desired de-noised image, the transformation operation tends to distort the structure of the image contents to a certain extent. We propose an improved algorithm for image enhancement and de-noising using the camera’s raw image data by employing a deep U-Net generator. The network is trained in an end-to-end manner on a large training set with suitable loss functions. To preserve the image content structures at a higher resolution compared to the existing approaches, we make use of an edge loss term in addition to PSNR loss and structural similarity loss during the training phase. Qualitative and quantitative results in terms of PSNR and SSIM values emphasize the effectiveness of our approach.

scholarly journals Depth Data Denoising in Optical Laser Based Sensors for Metal Sheet Flatness Measurement: A Deep Learning Approach

Sensors ◽  
2021 ◽  
Vol 21 (21) ◽  
pp. 7024
Author(s):  
Marcos Alonso ◽  
Daniel Maestro ◽  
Alberto Izaguirre ◽  
Imanol Andonegui ◽  
Manuel Graña
Keyword(s):  
Deep Learning ◽  
Optical Sensors ◽  
Optical Measurement ◽  
Image Data ◽  
Real Data ◽  
Noise Removal ◽  
Metal Sheet ◽  
Superior Performance ◽  
Range Image ◽  
Range Images

Surface flatness assessment is necessary for quality control of metal sheets manufactured from steel coils by roll leveling and cutting. Mechanical-contact-based flatness sensors are being replaced by modern laser-based optical sensors that deliver accurate and dense reconstruction of metal sheet surfaces for flatness index computation. However, the surface range images captured by these optical sensors are corrupted by very specific kinds of noise due to vibrations caused by mechanical processes like degreasing, cleaning, polishing, shearing, and transporting roll systems. Therefore, high-quality flatness optical measurement systems strongly depend on the quality of image denoising methods applied to extract the true surface height image. This paper presents a deep learning architecture for removing these specific kinds of noise from the range images obtained by a laser based range sensor installed in a rolling and shearing line, in order to allow accurate flatness measurements from the clean range images. The proposed convolutional blind residual denoising network (CBRDNet) is composed of a noise estimation module and a noise removal module implemented by specific adaptation of semantic convolutional neural networks. The CBRDNet is validated on both synthetic and real noisy range image data that exhibit the most critical kinds of noise that arise throughout the metal sheet production process. Real data were obtained from a single laser line triangulation flatness sensor installed in a roll leveling and cut to length line. Computational experiments over both synthetic and real datasets clearly demonstrate that CBRDNet achieves superior performance in comparison to traditional 1D and 2D filtering methods, and state-of-the-art CNN-based denoising techniques. The experimental validation results show a reduction in error than can be up to 15% relative to solutions based on traditional 1D and 2D filtering methods and between 10% and 3% relative to the other deep learning denoising architectures recently reported in the literature.

Image Denoising Based on Membrane Energy Filter

2011 ◽  
Vol 63-64 ◽  
pp. 345-349
Author(s):  
Ya Wei Li
Keyword(s):  
Image Denoising ◽  
Image Data ◽  
Rational Interpolation ◽  
Noise Removal ◽  
Experimental Methodology ◽  
Noise Point ◽  
Membrane Energy ◽  
Classical Image ◽  
Image Detail ◽  
Traditional Image

The traditional image denoising methods could effectively remove the noise. But the useful information in the detail abundant areas would be thrown off, and the edge appears mistiness. The search for efficient image denoising methods is still a valid challenge in image processing. This paper proposes a method for removing noise with keeping image detail in smoothness areas and detail abundant areas. The main focus of this paper is to define a general mathematical and experimental methodology to compare and classify classical image denoising algorithms. A bivariate rational interpolation with parameters is used in the algorithm. An interpolation surface is constructed using an image data as the interpolation data. According to the maximum and minimum membrane energy value of the interpolation surface, the noise pixel is detected. If it is a noise point, the value is replaced by the rank-ordered mean of the filter window and the membrane energy during noise removal. The experimental results demonstrate that the proposed method outperforms other conventional methods and recently proposed methods in reducing noise and retaining details.

Nonlinear Blind Source Separation Using GA Optimized RBF-ICA and its Application to the Image Noise Removal

2011 ◽  
Vol 393-395 ◽  
pp. 205-208 ◽  
Author(s):  
Xue Mei Wang
Keyword(s):  
Neural Network ◽  
Blind Source Separation ◽  
Rbf Neural Network ◽  
Source Separation ◽  
Image Data ◽  
Noise Removal ◽  
Independent Component ◽  
Data Set ◽  
Mixing Matrix ◽  
Ica Algorithm

Today the blind source separation (BSS) algorithms are widely used to separate independent components in a data set based on its statistical properties. Especially in image data applications, the independent component analysis (ICA) based BSS procedure for image pre-processing has been successfully applied for independent component extraction in order to remove the noise signals mixed into the image data. The contribution of this paper refers to the development of a nonlinear BSS method using the radial basis function (RBF) neural network based ICA algorithm, which was built by adopted some modifications in the linear ICA model. Moreover, genetic algorithm (GA) was used to optimize the RBF neural network to obtain satisfactory nonlinear solve of the nonlinear mixing matrix. In the experiments of this work, the GA optimized nonlinear ICA method and other ICA models were applied for image de-noising. A comparative analysis has showed satisfactory and effective image de-noising results obtained by the presented method.

scholarly journals Fourier Domain Anomaly Detection and Spectral Fusion for Stripe Noise Removal of TIR Imagery

2020 ◽  
Vol 12 (22) ◽  
pp. 3714
Author(s):  
Qingjie Zeng ◽  
Hanlin Qin ◽  
Xiang Yan ◽  
Tingwu Yang
Keyword(s):  
Remote Sensing ◽  
Anomaly Detection ◽  
Real Life ◽  
Image Data ◽  
Noise Removal ◽  
Fourier Domain ◽  
Transform Domain ◽  
Noise Pattern ◽  
Original Spectrum ◽  
Weight Map

Stripe noise is a common and unwelcome noise pattern in various thermal infrared (TIR) image data including conventional TIR images and remote sensing TIR spectral images. Most existing stripe noise removal (destriping) methods are often difficult to keep a good and robust efficacy in dealing with the real-life complex noise cases. In this paper, based on the intrinsic spectral properties of TIR images and stripe noise, we propose a novel two-stage transform domain destriping method called Fourier domain anomaly detection and spectral fusion (ADSF). Considering the principal frequencies polluted by stripe noise as outliers in the statistical spectrum of TIR images, our naive idea is first to detect the potential anomalies and then correct them effectively in the Fourier domain to reconstruct a desired destriping result. More specifically, anomaly detection for stripe frequencies is achieved through a regional comparison between the original spectrum and the expected spectrum that statistically follows a generalized Laplacian regression model, and then an anomaly weight map is generated accordingly. In the correction stage, we propose a guidance-image-based spectrum fusion strategy, which integrates the original spectrum and the spectrum of a guidance image via the anomaly weight map. The final reconstruction result not only has no stripe noise but also maintains image structures and details well. Extensive real experiments are performed on conventional TIR images and remote sensing spectral images, respectively. The qualitative and quantitative assessment results demonstrate the superior effectiveness and strong robustness of the proposed method.

scholarly journals Rough or Noisy? Metrics for Noise Estimation in SfM Reconstructions

Sensors ◽  
2020 ◽  
Vol 20 (19) ◽  
pp. 5725
Author(s):  
Ivan Nikolov ◽  
Claus Madsen
Keyword(s):  
Image Data ◽  
Ground Truth ◽  
Noise Removal ◽  
Difficult Problem ◽  
Noise Estimation ◽  
Real Surface ◽  
Reconstructed Surfaces ◽  
Noise Assessment ◽  
Global Smoothing ◽  
Statistical Metrics

Structure from Motion (SfM) can produce highly detailed 3D reconstructions, but distinguishing real surface roughness from reconstruction noise and geometric inaccuracies has always been a difficult problem to solve. Existing SfM commercial solutions achieve noise removal by a combination of aggressive global smoothing and the reconstructed texture for smaller details, which is a subpar solution when the results are used for surface inspection. Other noise estimation and removal algorithms do not take advantage of all the additional data connected with SfM. We propose a number of geometrical and statistical metrics for noise assessment, based on both the reconstructed object and the capturing camera setup. We test the correlation of each of the metrics to the presence of noise on reconstructed surfaces and demonstrate that classical supervised learning methods, trained with these metrics can be used to distinguish between noise and roughness with an accuracy above 85%, with an additional 5–6% performance coming from the capturing setup metrics. Our proposed solution can easily be integrated into existing SfM workflows as it does not require more image data or additional sensors. Finally, as part of the testing we create an image dataset for SfM from a number of objects with varying shapes and sizes, which are available online together with ground truth annotations.

Intelligent computational 3-D microscopy: New results in segmentation, noise removal, cell counting, and automated neuron tracing

1993 ◽  
Vol 51 ◽  
pp. 160-161
Author(s):  
Badrinath Roysam ◽  
Anoop K. Bhattacharjya ◽  
Hakan Ancin ◽  
Andrew R. Cohen ◽  
Robert W. Mackin ◽  
...  
Keyword(s):  
Laser Scanning ◽  
Image Interpretation ◽  
Image Data ◽  
Noise Removal ◽  
Laser Scanning Confocal Microscopy ◽  
Cell Counting ◽  
Scanning Confocal Microscopy ◽  
Neuron Tracing ◽  
Large Sets ◽  
Automated Pattern Recognition

This paper presents a review of recent advances made by us in the broad area of intelligent computational 3-D microscopy. This research involves the development and application of automated pattern recognition, computer vision and image interpretation algorithms to problems in 3-D biological microscopy. The broad goal of this work is to develop computational tools for rapid, unsupervised, flexible, quantitative and unbiased analysis of large sets of complex 3-D images of thick biological specimens. The imaging modalities used in this work have thus far involved laser-scanning confocal microscopy (LSCM) and 3-D brightfield microscopy. The specimens of interest have been thick slices of brain tissue, large selectively-stained neurons, indocyanin green and fluorescein retinal angiograms, and duck overlapping cell clusters in cytological preparations such as Papanicolaou smears.Image segmentation is an essential first step for most forms of analysis. This task is straightforward when the image data is isotropic, has uniform high contrast, and is free of noise.

Dynamical Scattering in Crystalline Biological Specimens. I. Criteria of Validity for Scattering Approximations

1975 ◽  
Vol 33 ◽  
pp. 192-193
Author(s):  
Robert M. Glaeser ◽  
Bing K. Jap
Keyword(s):  
Biological Sciences ◽  
Electron Microscopy ◽  
Electron Microscope ◽  
Electron Diffraction ◽  
Born Approximation ◽  
Materials Science ◽  
Image Data ◽  
Dynamical Effect ◽  
Crystallographic Structure ◽  
Electron Microscope Image

The dynamical scattering effect, which can be described as the failure of the first Born approximation, is perhaps the most important factor that has prevented the widespread use of electron diffraction intensities for crystallographic structure determination. It would seem to be quite certain that dynamical effects will also interfere with structure analysis based upon electron microscope image data, whenever the dynamical effect seriously perturbs the diffracted wave. While it is normally taken for granted that the dynamical effect must be taken into consideration in materials science applications of electron microscopy, very little attention has been given to this problem in the biological sciences.

Application of an image management system in microscopy

1992 ◽  
Vol 50 (2) ◽  
pp. 1052-1053
Author(s):  
Richard S. Chemock
Keyword(s):  
Data Storage ◽  
Cost Savings ◽  
Image Data ◽  
Analytical Techniques ◽  
Operational Mode ◽  
Fine Grained ◽  
Image Recording ◽  
Primary Output ◽  
Capacity Data ◽  
Image Management System

One of the most common tasks in a typical analysis lab is the recording of images. Many analytical techniques (TEM, SEM, and metallography for example) produce images as their primary output. Until recently, the most common method of recording images was by using film. Current PS/2R systems offer very large capacity data storage devices and high resolution displays, making it practical to work with analytical images on PS/2s, thereby sidestepping the traditional film and darkroom steps. This change in operational mode offers many benefits: cost savings, throughput, archiving and searching capabilities as well as direct incorporation of the image data into reports.The conventional way to record images involves film, either sheet film (with its associated wet chemistry) for TEM or PolaroidR film for SEM and light microscopy. Although film is inconvenient, it does have the highest quality of all available image recording techniques. The fine grained film used for TEM has a resolution that would exceed a 4096x4096x16 bit digital image.

Sign in / Sign up

Export Citation Format

Share Document