An Improved Image Denoising Method Based on Wavelet Thresholding

VisuShrink, ModineighShrink and NeighShrink are efficient image denoising algorithms based on the discrete wavelet transform (DWT). These methods have disadvantage of using a suboptimal universal threshold and identical neighbouring window size in all wavelet subbands. In this paper, an improved method is proposed, that determines a threshold as well as neighbouring window size for every subband using its lengths. Our experimental results illustrate that the proposed approach is better than the existing ones, i.e., NeighShrink, ModineighShrink and VisuShrink in terms of peak signal-to-noise ratio (PSNR) i.e. visual quality of the image.

Research on Destriping Based on Rough Set

In order to remove the stripe noises in cotton foreign fiber images by line scanning camera collected, in multi threshold segmentation of rough set, every region’s color is instead of the statistics color of the region. This method can retain the detail information of original image as far as possible, and do well in the stripe noise removal. The roughness of rough set was calculated respectively using directional diagram, Canny operator and Sobel operator. Comparing the three methods, the results indicate that the Canny operator keeps the more details of image, and directional diagram and Sobel operator have the better effects on denoising.

Junk band recovery for hyperspectral image based on curvelet transform

Under consideration that the profiles of bands at close wavelengths are quite similar and the curvelets are good at capturing profiles, a junk band recovery algorithm for hyperspectral data based on curvelet transform is proposed. Both the noisy bands and the noise-free bands are transformed via curvelet band by band. The high frequency coefficients in junk bands are replaced with linear interpolation of the high frequency coefficients in noise-flee bands, and the low frequency coefficients remain the same to keep the main spectral characteristics from being distorted. Jutak bands then are recovered after the inverse curvelet transform. The performance of this method is tested on the hyperspectral data cube obtained by airborne visible/infrared imaging spectrometer （AVIRIS）. The experimental results show that the proposed method is superior to the traditional denoising method BayesShrink and the art-of-state Curvelet Shrinkage in both roots of mean square error （RMSE） and peak-signal-to-noise ratio （PSNR） of recovered bands.

Research on Denoising of Cryo-em Images Based on Deep Learning

Cryo-em(Cryogenic electron microscopy)is a technology this can build bio-macromolecule of three-dimensional structure.Under the condition of now,the projection image of the biological macromolecule which is collected by the Cryo-em technology that the contrast is low,the signal to noise is low,image blurring,and not easy to distinguish single particle from background,the corresponding processing technology is lagging behind.Therefore,make Cryo-em image denoising useful,and maintaining bio-macromolecule of contour or signal of function-construct improve Cryo-em image quality or resolution of Cryo-em three-dimensional structure have important effect.This paper researched a denoising function base on GANs(generative adversarial networks),purpose an improved discriminant model base on Wasserstein distance and an improved image denoising model by add gray constraint.Our model turn discriminant model’s training process from binary classification’s training process into regression task training process,it make GANs in training process more stable,more reasonable parameter passing.Meantime,we also propose an improved generative model by add gray constraint.The experimental results show that our model can increase the peak signal-to-noise ratio of the Cryo-em simulation image by 10.3 dB and improve SSIM(Structural Similarity Index)of the denoised image results by 0.43.Compared with traditional image denoising algorithms such as BM3D(Block Matching 3D),our model can better save the model structure and the vein signal in the original image and the operation speed is faster.

Practical Blind Image Denoising via Swin-Conv-UNet and Data Synthesis

While recent years have witnessed a dramatic upsurge of exploiting deep neural networks toward solving image denoising,existing methods mostly rely on simple noise assumptions,such as additive white Gaussian noise(AWGN),JPEG compression noise and camera sensor noise,and a general-purpose blind denoising method for real images remains unsolved.In this paper,we attempt to solve this problem from the perspective of network architecture design and training data synthesis.Specifically,for the network architecture design,we propose a swin-conv block to incorporate the local modeling ability of residual convolutional layer and non-local modeling ability of swin transformer block,and then plug it as the main building block into the widely-used image-to-image translation UNet architecture.For the training data synthesis,we design a practical noise degradation model which takes into consideration different kinds of noise(including Gaussian,Poisson,speckle,JPEG compression,and processed camera sensor noises)and resizing,and also involves a random shuffle strategy and a double degradation strategy.Extensive experiments on AGWN removal and real image denoising demonstrate that the new network architecture design achieves state-of-the-art performance and the new degradation model can help to significantly improve the practicability.We believe our work can provide useful insights into current denoising research.The source code is available at https://github.com/cszn/SCUNet.

An 8.12 μW wavelet denoising chip for PPG detection and portable heart rate monitoring in 0.18 μm CMOS

A low power wavelet denoising chip for photoplethysmography （PPG） detection and portable heart rate monitoring is presented. To eliminate noise and improve detection accuracy, Harr wavelet （HWT） is chosen as the processing tool. An optimized finite impulse response structure is proposed to lower the computational complexity of proposed algorithm, which is benefit for reducing the power consumption of proposed chip. The modulus max- ima pair location module is design to accurately locate the PPG peaks. A clock control unit is designed to further reduce the power consumption of the proposed chip. Fabricated with the 0.18μm N-well CMOS 1P6M technol- ogy, the power consumption of proposed chip is only 8.12μW in 1 V voltage supply. Validated with PPG signals in multiparameter intelligent monitoring in intensive care databases and signals acquired by the wrist photoelectric volume detection front end, the proposed chip can accurately detect PPG signals. The average sensitivity and positive prediction are 99.91% and 100%, respectively.

A statistical learning based image denoising approach

The image denoising is a very basic but important issue in the field of image procession. Most of the existing methods addressing this issue only show desirable performance when the image complies with their underlying assumptions. Especially, when there is more than one kind of noises, most of the existing methods may fail to dispose the corresponding image. To address this problem, we propose a two-step image denoising method motivated by the statistical learning theory. Under the proposed framework, the type and variance of noise are estimated with support vector machine （SVM） first, and then this information is employed in the proposed denoising algorithm to further improve its denoising performance. Finally, comparative study is constructed to demonstrate the advantages and effectiveness of the proposed method.