Comparison of PDE-based nonlinear diffusion approaches for image enhancement and denoising in optical coherence tomography

A comparison between two nonlinear diffusion methods for denoising OCT images is performed. Specifically, we compare and contrast the performance of the traditional nonlinear Perona-Malik filter with a complex diffusion filter that has been recently introduced by Gilboa et al.. The complex diffusion approach based on the generalization of the nonlinear scale space to the complex domain by combining the diffusion and the free Schridinger equation is evaluated on synthetic images and also on representative OCT images at various noise levels. The performance improvement over the traditional nonlinear Perona-Malik filter is quantified in terms of noise suppression, image structural preservation and visual quality. An average signal-to-noise ratio (SNR) improvement of about 2.5 times and an average contrast to noise ratio (CNR) improvement of 49% was obtained while mean structure similarity (MSSIM) was practically not degraded after denoising. The nonlinear complex diffusion filtering can be applied with success to many OCT imaging applications. In summary, the numerical values of the image quality metrics along with the qualitative analysis results indicated the good feature preservation performance of the complex diffusion process, as desired for better diagnosis in medical imaging processing.     

Nonlinear noise cancellation for image with adaptive neuro-fuzzy inference systems

The adaptive neural-fuzzy inference system (ANFIS) algorithm is proposed for nonlinear noise cancellation of images. The quality in terms of mean square error of image restoration using the proposed ANFIS is over 65 times better for Gaussian noise and 500 times better for salt and pepper noise than using conventional filtering systems.     

Anisotropic diffusion for noise removal of band pass signals

A noise removal method for band pass signals based on the anisotropic diffusion algorithm originally put forward by Perona and Malik is proposed in this paper. The anisotropic smoothing algorithm proposed here is for band pass signals modulated with a constant carrier frequency. A partial differential equation to smooth band pass noisy signals is derived. The propagator of this differential equation is also analytically calculated in this paper. An appropriate linear operator is then considered here for such band pass signals to form an anisotropic diffusion algorithm. The algorithm proposed here demonstrates better performance for band pass noisy signals containing discontinuities in comparison with the traditional Perona-Malik (PM) algorithm and is robust in the presence of excessive noise with SNR less than unity.     

嵌入固有模态函数的各向异性扩散方程用于图像降噪

该文利用经验模态分解技术对图像进行分解,获得表示图像不同频率属性的各个固有模态函数分量,并将代表图像高频信息和次高频信息的固有模态函数嵌入到Perona-Malik模型中。改进后的模型不仅在对高斯噪声降噪时优于原Perona-Malik模型,而且对椒盐噪声也能较好地去除。     

图像中的噪声及其复原方法的比较研究

图像降噪作为图像预处理的必要步骤,在图像处理中有着重要的意义。由于在获取图像的过程中,成像系统受到各种因素的影响,会导致图像质量的降低,图像复原技术就是趋向于将退化模型化并用相反的处理来恢复原图像。本文针对引起图像质量退化的噪声因素进行了分析研究,并对采集图像中常见的高斯白噪声和椒盐噪声进行了降噪研究。本文重点运用图像降噪方法中的中值滤波算法和小波阈值降噪方法进行了仿真实验,并分别利用这两种算法对同一幅噪声图像降噪进行效果比较并得到结论。     

Removal of impulse noise from color images based on the localized image characteristics and noise level

The proposed method is a switching vector median filter that utilizes suitable noise detection and restoration algorithms for different impulse noise densities. It works in a multistage manner with an enhancing accuracy of noise detection in each successive stage. The processing window is initially categorized into a non-edge-window and an edge-window, depending on the scale of abrupt transitions. The non-edge-window is further verified to be a smooth-window or a disorder-window, whereas the edge-window is again confirmed to be noise-free or a noisy edge-window. The proposed method is simulated on a variety of medical images and other standard test images to prove its efficiency in detection of noise and restoration.     

Phase noise in LC oscillators

Analytical methods for the phase-noise analysis of LC-tuned oscillators are presented. The fundamental assumption used in the theoretical model is that an oscillator acts as a large-signal LC-tuned amplifier for purposes of noise analysis. This approach allows us to derive closed-form expressions for the close-to-carrier spectral density of the output noise, and to estimate the phase-noise performance of an oscillator from circuit parameters using hand analysis. The emphasis is on an engineering approach intended to facilitate rapid estimation of oscillator phase noise. Theoretical predictions are compared with results of circuit simulations using a nonlinear phase-noise simulator. The analytical results are in good agreement with simulations for weakly nonlinear oscillators. Complete nonlinear simulations are necessary to accurately predict phase noise in oscillators operating in a strongly nonlinear regime. To confirm the validity of the nonlinear phase-noise models implemented in the simulator, simulation results are compared with measurements of phase noise in a practical Colpitts oscillator, where we find good agreement between simulations and measurements     

Using Anfis with circular polygons for impulsive noise suppression from highly distorted images

In this paper, a novel approach is presented to the restoration of images corrupted by impulsive noise (IN), with a new nonlinear IN suppression filter, entitled circular polygons based adaptive-fuzzy filter (CF). The proposed filter is based on statistical impulse detection and nonlinear filtering which uses adaptive-network-based fuzzy inference system (Anfis) as a missed data interpolant over the circular polygons and provides estimates for the original intensity values of corrupted pixels. Impulse detection is realized by using the chi-square based goodness-of-fit test, which yields a decision about the impulsivity of each pixel. Extensive simulations were realized to demonstrate the capability of CF and they reveal that the proposed filter achieves a better performance than the other filters mentioned in this paper in the cases of being effective in noise suppression and detail preservation, also when the images are highly corrupted by IN.     

Nonlinear multiscale filtering

In this article, we give an overview of scale-spaces and their application to noise suppression and segmentation of 1-D signals and 2-D images. Several prototypical problems serve as our motivation. We review several scale-spaces (linear Gaussian, Perona-Malik, and SIDE-stabilized inverse diffusion equation) and discuss their advantages and shortcomings. We describe our previous work and argue that a very simple nonlinear scale-space leads to a fast estimation algorithm which produces accurate segmentations and estimates of signals and images.     

Deterministic-chaotic variably coloured noise

By coupling several strange attractors (i.e. unstable nonlinear loops, or nonlinear recursive finite difference equations), the hardware or software needed for generating variably coloured noise can be simplified. It is noteworthy that this noise tends to take on the ubiquitous excess low-frequency form, characteristic of nonlinear devices, as the number of attractors increases.     

Using uncorrupted neighborhoods of the pixels for impulsive noise suppression with ANFIS.

In this paper, a novel adaptive network-based fuzzy inference system (ANFIS)-based filter, ABF, is presented for the restoration of images corrupted by impulsive noise (IN). The ABF is performed in two steps. In the first step, impulse detection is realized by using statistical tools. In the second step, a nonlinear filtering scheme based on ANFIS is performed for only the corrupted pixels detected in the first step. To demonstrate the effectivity of ABF at the removal of high-level IN, extensive simulations were realized for ABF and nine different comparison filters. Empirical results indicate that the proposed filter achieves a better performance than the comparison filters in terms of noise suppression and detail preservation, even when the images are highly corrupted by IN.     

一种视频监控图像条纹噪声去除算法

针对视频图像中图像传感器故障引起的等间距横纹噪声去除问题,提出了基于陷波滤波器的去除条纹噪声算法;首先将视频条纹图进行傅里叶变换,通过频域累积分布函数映射法构造自适应滤波器,用该滤波器对变换后的条纹图像进行滤波,最后用Laplacian算子进行锐化处理得到最终去噪声图像。实验结果表明,本算法适合用于去除有图像传感器故障引起的等间距横纹噪声。     

Adaptive scheme for elimination of broadband noise and impulsivedisturbances from AR and ARMA signals

The problem of elimination of measurement noise and impulsive disturbances from autoregressive and autoregressive moving average signals is considered. It is shown that the task of simultaneous detection/tracking/restoration can be stated as a nonlinear filtering problem and solved using the theory of extended Kalman filter. Numerical tests carried out for real audio signals corrupted by both real and artificially generated disturbances confirm very good properties of the proposed algorithm     

An advanced low-frequency noise model of GaInP-GaAs HBT for accurate prediction of phase noise in oscillators

We present a new low-frequency noise model of a GaInP-GaAs HBT and the associated extraction process from measurements. Specific measurements enable us to locate the two dominant low-frequency noise sources. Their spectral densities extraction as a function of the emitter bias current is then performed and a normalized scalable model is deduced. The cyclostationarity of the low-frequency noise sources is justified. The whole noise model including the shot noise source is implemented in the nonlinear HBT model used in the United Monolithic Semiconductors foundry. In order to verify the validity of the scalable noise model, several voltage-controlled oscillators with different center frequencies and tuning bandwidth have been designed and processed. Comparisons between the predicted performances and experimental results show an excellent agreement and validate the proposed low-frequency noise modeling of multifinger HBTs.     

基于局部最小相关系数的自适应扩散迭代次数研究

针对典型Perona-Malik各向异性扩散去噪模型中采用固定迭代次数的不足,提出了一种自适应迭代次数确定算法,通过去噪后图像与噪声的局部最小相关系数来自动确定扩散迭代次数.实验结果表明:本文算法比传统算法去噪效果更好,同时也很好地保持了图像的边缘细节,且具有一定的通用性.     

Memoryless discrete-time detection of a constant signal in m-dependent noise

The problem of designing memoryless detectors for known signals in stationary m-dependent noise processes is considered. Applying the criterion of asymptotic relative efficiency, the optimal such detector is shown to be characterized by the solution to a Fredholm integral equation whose kernel depends only on the second-order probability distributions of the noise. General expressions are derived for this solution and for the asymptotic efficiency of the optimal detector relative to other memoryless detectors. To illustrate the analysis, specific results are given for the particular case where the noise process is derived by memoryless nonlinear transformation of a Gaussian process. In addition, an extension of the analytical results to the more general case ofphi-mixing noise processes is discussed.     

基于小波神经网络的非线性噪声对消

利用小波神经网络作为噪声对消滤波器,实现了参考噪声与干扰噪声呈非线性相关条件下的噪声对消。在参考噪声与干扰噪声非线性相关时,传统的横向滤波器效果不理想,利用小波神经网络的非线性特性,可更好的解决非线性噪声条件下的噪声对消问题。计算机仿真结果证明,小波神经网络噪声对消在非线性噪声条件可有效提高信噪比增益。     

基于噪声分析和稀疏正则化的图像盲复原方法

在海事搜救过程中,机载红外相机拍摄的红外图像由于直升机振动、气流扰动、高速飞行以及红外相机摆扫等因素,严重影响图像质量.根据直升机载红外相机成像特点,提出了一种基于噪声分析和稀疏正则化的图像盲复原方法.该方法首先分析了成像过程中的噪声分布,并对噪声进行预处理,再根据稀疏表达理论,用图像边缘的稀疏先验信息指导点扩散函数复原,接着通过非盲复原方法得到目标图像,将目标图像作为下一次迭代的输入图像,如此循环迭代得到清晰图像.最后,对仿真模糊图像和实拍模糊图像进行了复原实验.实验结果表明这种方法能有效改善图像质量,并且在处理实拍运动模糊图像时,相比其他复原方法效果更好.     

Analysis of noise up-conversion in microwave field-effecttransistor oscillators

The conversion process of the low frequency noise into phase noise in field-effect transistors (FET) oscillators is investigated. First, an evaluation of the baseband noise contribution to the oscillator phase noise is provided from the analysis of the baseband noise and the frequency noise spectra. A distinction is made within the different components of the low frequency noise contributions to close-in carrier phase noise. Next, the frequency noise of the oscillator circuit is analyzed in terms of the FET's low frequency noise multiplied by the oscillator's pushing factor. Though this product usually provides a good evaluation of the phase noise, experimental results presented here show the inaccuracy of this method at particular gate bias voltages where the pushing factor decreases to zero. To account for these observations, a new nonlinear FET model involving at least two noise sources distributed along the channel is proposed     

Dual learning based compression noise reduction in the texture domain

Compression noise reduction is similar to the super-resolution problem in terms of the restoration of lost high-frequency information. Because learning-based approaches have proven successful in the past in terms of addressing the super-resolution problem, we focus on a learning-based technique for compressed image denoising. In this process, it is important to search for the exact prior in a training set. The proposed method utilizes two different databases (i.e., a noisy and a denoised database), which work together in a complementary way. The denoised images from the dual databases are combined into a final denoised one. Additionally, the input noisy image is decomposed into structure and texture components, and only the latter is denoised because most noise tends to exist within the texture component. Experimental results show that the proposed method can reduce compression noise while reconstructing the original information that was lost in the compression process, especially for texture regions.