基于结构张量的Non-Local Means去噪算法研究

非局部平均是当前一种新兴而有效的图像去噪方法。为了能充分利用数字图像局部几何结构的自相似性,同时由于结构张量可有效刻画数字图像的局部几何结构特征,进而提出了基于结构张量相似性度量的非局部平均去噪算法。实验结果验证了该算法抑制噪声的有效性,同时能很好地保持边缘等细节特征,峰值信噪比得到有效提高。

Review of local mean decomposition and its application in fault diagnosis of rotating machinery

Rotating machinery is widely used in the industry.They are vulnerable to many kinds of damages especially for those working under tough and time-varying operation conditions.Early detection of these damages is important,otherwise,they may lead to large economic loss even a catastrophe.Many signal processing methods have been developed for fault diagnosis of the rotating machinery.Local mean decomposition(LMD)is an adaptive mode decomposition method that can decompose a complicated signal into a series of mono-components,namely product functions(PFs).In recent years,many researchers have adopted LMD in fault detection and diagnosis of rotating machines.We give a comprehensive review of LMD in fault detection and diagnosis of rotating machines.First,the LMD is described.The advantages,disadvantages and some improved LMD methods are presented.Then,a comprehensive review on applications of LMD in fault diagnosis of the rotating machinery is given.The review is divided into four parts:fault diagnosis of gears,fault diagnosis of rotors,fault diagnosis of bearings,and other LMD applications.In each of these four parts,a review is given to applications applying the LMD,improved LMD,and LMD-based combination methods,respectively.We give a summary of this review and some future potential topics at the end.

Fast Non-Local Means Algorithm Based on Krawtchouk Moments

Non-local means(NLM)method is a state-of-the-art denoising algorithm, which replaces each pixel with a weighted average of all the pixels in the image. However, the huge computational complexity makes it impractical for real applications. Thus, a fast non-local means algorithm based on Krawtchouk moments is proposed to improve the denoising performance and reduce the computing time. Krawtchouk moments of each image patch are calculated and used in the subsequent similarity measure in order to perform a weighted averaging. Instead of computing the Euclidean distance of two image patches, the similarity measure is obtained by low-order Krawtchouk moments, which can reduce a lot of computational complexity. Since Krawtchouk moments can extract local features and have a good antinoise ability, they can classify the useful information out of noise and provide an accurate similarity measure. Detailed experiments demonstrate that the proposed method outperforms the original NLM method and other moment-based methods according to a comprehensive consideration on subjective visual quality, method noise, peak signal to noise ratio(PSNR), structural similarity(SSIM) index and computing time. Most importantly, the proposed method is around 35 times faster than the original NLM method.

Improved Non-Local Means Algorithm for Image Denoising

Image denoising technology is one of the forelands in the field of computer graphic and computer vision. Non-local means method is one of the great performing methods which arouse tremendous research. In this paper, an improved weighted non-local means algorithm for image denoising is proposed. The non-local means denoising method replaces each pixel by the weighted average of pixels with the surrounding neighborhoods. The proposed method evaluates on testing images with various levels noise. Experimental results show that the algorithm improves the denoising performance.

Two Modifications of Weight Calculation of the Non-Local Means Denoising Method

The non-local means (NLM) denoising method replaces each pixel by the weighted average of pixels with the sur-rounding neighborhoods. In this paper we employ a cosine weighting function instead of the original exponential func-tion to improve the efficiency of the NLM denoising method. The cosine function outperforms in the high level noise more than low level noise. To increase the performance more in the low level noise we calculate the neighborhood si-milarity weights in a lower-dimensional subspace using singular value decomposition (SVD). Experimental compari-sons between the proposed modifications against the original NLM algorithm demonstrate its superior denoising per-formance in terms of peak signal to noise ratio (PSNR) and histogram, using various test images corrupted by additive white Gaussian noise (AWGN).

启发式k-means聚类算法的改进研究

启发式k-means聚类算法通过在k-means第一次迭代后查看附近的集群来预测每个数据点可能会被划分到的集群子集,有效地加快了算法的运行速度。但由于启发式算法存在随机选择初始聚类中心以及无法有效识别数据集中离群点的缺陷,导致聚类结果的误差平方和较大并且轮廓系数偏小。针对这一问题,提出了CHk-means算法,该算法引入仔细播种方法,克服了启发式k-means算法随机选择初始聚类中心带来的局部最优解问题;该算法引入局部异常因子LOF算法对离群点进行检测,降低了离群点数据对聚类结果的影响。在多个数据集上对3种算法进行对比试验,结果表明CHk-means算法可有效降低聚类结果的误差平方和,增强聚类的轮廓系数,使聚类质量得到明显改善。

A Robust and Fast Non-Local Means Algorithm for Image Denoising

In the paper, we propose a robust and fast image denoising method. The approach integrates both Non- Local means algorithm and Laplacian Pyramid. Given an image to be denoised, we first decompose it into Laplacian pyramid. Exploiting the redundancy property of Laplacian pyramid, we then perform non-local means on every level image of Laplacian pyramid. Essentially, we use the similarity of image features in Laplacian pyramid to act as weight to denoise image. Since the features extracted in Laplacian pyramid are localized in spatial position and scale, they are much more able to describe image, and computing the similarity between them is more reasonable and more robust. Also, based on the efficient Summed Square Image （SSI） scheme and Fast Fourier Transform （FFT）, we present an accelerating algorithm to break the bottleneck of non-local means algorithm - similarity computation of compare windows. After speedup, our algorithm is fifty times faster than original non-local means algorithm. Experiments demonstrated the effectiveness of our algorithm.

Feature Extraction of Bearing Vibration Signals Using Second Generation Wavelet and Spline-Based Local Mean Decomposition

In order to extract the fault feature frequency of weak bearing signals,we put forward a local mean decomposition(LMD)method combining with the second generation wavelet transform.After performing the second generation wavelet denoising,the spline-based LMD is used to decompose the high-frequency detail signals of the second generation wavelet signals into a number of production functions(PFs).Power spectrum analysis is applied to the PFs to detect bearing fault information and identify the fault patterns.Application in inner and outer race fault diagnosis of rolling bearing shows that the method can extract the vibration features of rolling bearing fault.This method is suitable for extracting the fault characteristics of the weak fault signals in strong noise.

Weak characteristic information extraction from early fault of wind turbine generator gearboxKeywords wind turbine generator gearbox, B-singular value decomposition, local mean decomposition, weak characteristic information extraction, early fault warning

Given the weak early degradation characteristic information during early fault evolution in gearbox of wind turbine generator, traditional singular value decomposition （SVD）-based denoising may result in loss of useful information. A weak characteristic information extraction based on μ-SVD and local mean decomposition （LMD） is developed to address this problem. The basic principle of the method is as follows： Determine the denoising order based on cumulative contribution rate, perform signal reconstruction, extract and subject the noisy part of signal to LMD and μ-SVD denoising, and obtain denoised signal through superposition. Experimental results show that this method can significantly weaken signal noise, effectively extract the weak characteristic information of early fault, and facilitate the early fault warning and dynamic predictive maintenance.

The Algorithms about Fast Non-local Means Based Image Denoising

Image denoising is still a challenge of image processing. Buades et al. proposed a nonlocal means （NL-means） approach. This method had a remarkable denoising results at high expense of computational cost. In this paper, We compared several fast non-local means methods, and proposed a new fast algorithm. Numerical experiments showed that our algorithm considerably reduced the computational cost, and obtained visually pleasant images.

基于Walsh-Hadamard投影的快速Nonlocal-Means图像去噪

针对Nonlocal-means图像去噪算法计算强度较高的问题,提出了一种投影加速算法。分析了Nonlo-cal-means算法的计算瓶颈,利用Walsh-Hadamard变换的快速运算和自递归特性,使用一组完备的Walsh-Hadamard变换将图像块投影到其张成的空间中;利用Walsh-Hadamard投影的能量集中特性和匹配计算过程中的拒绝策略,在Nonlocal-means去噪算法的图像块匹配计算中快速丢弃无法匹配的图像块。实验结果表明,该算法获得了较好的加速性能,且去噪效果没有受到影响。

A Two-Step Regularization Framework for Non-Local Means

As an effective patch-based denoising method, non-local means （NLM） method achieves favorable denoising performance over its local counterparts and has drawn wide attention in image processing community. The in, plementation of NLM can formally be decomposed into two sequential steps, i.e., computing the weights and using the weights to compute the weighted means. In the first step, the weights can be obtained by solving a regularized optimization. And in the second step, the means can be obtained by solving a weighted least squares problem. Motivated by such observations, we establish a two-step regularization framework for NLM in this paper. Meanwhile, using the fl-amework, we reinterpret several non-local filters in the unified view. Further, taking the framework as a design platform, we develop a novel non-local median filter for removing salt-pepper noise with encouraging experimental results.

Local electron mean energy profile of positive primary streamer discharge with pin-plate electrodes in oxygen nitrogen mixtures

Local electron mean energy （LEME） has a direct effect on the rates of collisional ionization of molecules and atoms by electrons. Electron-impact ionization plays an important role and is the main process for the production of charged particles in a primary streamer discharge. Detailed research on the LEME profile in a primary streamer discharge is extremely important for a comprehensive understanding of the local physical mechanism of a streamer. In this study, the LEME profile of the primary streamer discharge in oxygen-nitrogen mixtures with a pin-plate gap of 0.5 cm under an impulse voltage is investigated using a fluid model. The fluid model includes the electron mean energy density equation, as well as continuity equations for electrons and ions and Poisson＇s electric field equation. The study finds that, except in the initial stage of the primary streamer, the LEME in the primary streamer tip tends to increase as the oxygen-nitrogen mole ratio increases and the pressure decreases. When the primary streamer bridges the gap, the LEME in the primary streamer channel is smaller than the first ionization energies of oxygen and nitrogen. The LEME in the primary streamer channel then decreases as the oxygen-nitrogen mole ratio increases and the pressure increases. The LEME in the primary streamer tip is primarily dependent on the reduced electric field with mole ratios of oxygen-nitrogen given in the oxygen-nitrogen mixtures.

A powerful denoising method based on non-local means filter for cryo-electron microscopic images

Cryo-electron microscopic images of biological molecules usually have high noise and low contrast. It is essential to suppress noise and enhance contrast in order to recognize

基于K-means的GLOCAL改进算法

在多标签学习中,标签相关性是不可或缺的。当标签缺损只能观察到一部分标签时,很难判断标签之间的相关性。具有全局与局部标签相关性的多标签(GLOCAL)算法通过学习潜在标签和引入标签流行正则化项,同时利用全局和局部标签相关性来解决标签缺损问题。但是该算法在通过低秩分解学习潜在标签以及原始标签与潜在标签的关联性时,初始化的低秩矩阵是随机获取的,这导致该算法结果并不稳定。基于此,利用K-means算法对原始标签进行聚类,获得的聚类中心矩阵将能更好地表现出原始标签与潜在标签之间的相关性。实验结果表明,本文的算法是合理和有效的。

ON THE LOCALIZATION AND CONVERGENCE OF MULTIPLE FOURIER INTEGRAL BY BOCHNER-RIESZ MEANS

In this paper we consider lim _(R-) B_R^(f,x_0), in one case that f_x_0 (t) is a ABMV function on [0, ∞], and in another case that f∈L_(m-1)~1(R~) and x^k/~kf∈BV(R) when |k| = m-1 and f(x) = 0 when |x -x_0|<δ for some δ>0. Our theormes improve the results of Pan Wenjie ([1]).

融合最近邻矩阵与局部密度的自适应K-means聚类算法

针对传统K-means聚类算法对初始聚类中心和离群孤立点敏感的缺陷,以及现有引入密度概念优化的K-means算法均需要设置密度参数或阈值的缺点,提出一种融合最近邻矩阵与局部密度的自适应K-means聚类算法。受最邻近吸收原则与密度峰值原则启发,通过引入数据对象间的距离差异值构造邻近矩阵,根据邻近矩阵计算局部密度,不需要任何参数设置,采取最近邻矩阵与局部密度融合策略,自适应确定初始聚类中心数目和位置,同时完成非中心点的初分配。人工数据集和UCI数据集的实验测试,以及与传统K-means算法、基于离群点改进的K-means算法、基于密度改进的K-means算法的实验比较表明,提出的自适应K-means算法对人工数据集的孤立点免疫度较高,对UCI数据集具有更准确的聚类结果。

面向大型数据集的局部敏感哈希K−means算法

大型数据集高效处理策略是煤矿安全监测智能化、采掘智能化等煤矿智能化建设的关键支撑。针对K−means算法面对大型数据集时聚类高效性及准确性不足的问题,提出了一种基于局部敏感哈希(LSH)的高效K−means聚类算法。基于LSH对抽样过程进行优化,提出了数据组构建算法LSH−G,将大型数据集合理划分为子数据组,并对数据集中的噪声点进行有效删除;基于LSH−G算法优化密度偏差抽样(DBS)算法中的子数据组划分过程,提出了数据组抽样算法LSH−GD,使样本集能更真实地反映原始数据集的分布规律;在此基础上,通过K−means算法对生成的样本集进行聚类,实现较低时间复杂度情况下从大型数据集中高效挖掘有效数据。实验结果表明:由10个AND操作与8个OR操作组成的级联组合为最优级联组合,得到的类中心误差平方和(SSEC)最小;在人工数据集上,与基于多层随机抽样(M−SRS)的K−means算法、基于DBS的K−means算法及基于网格密度偏差抽样(G−DBS)的K−means算法相比,基于LSH−GD的K−means算法在聚类准确性方面的平均提升幅度分别为56.63%、54.59%及25.34%,在聚类高效性方面的平均提升幅度分别为27.26%、16.81%及7.07%;在UCI标准数据集上,基于LSH−GD的K−means聚类算法获得的SSEC与CPU消耗时间(CPU−C)均为最优。

基于局部均值分解与局部离群因子动力电池故障诊断

动力电池故障诊断是保证电动汽车正常运行的关键。提出一种基于局部均值分解和局部离群因子的动力电池故障诊断方法,用于电池组故障识别与定位。通过局部均值分解对电压信号预处理,并根据相关系数高低重构电压信号。进一步提取重构信号的峭度因子作为故障特征输入到局部离群因子算法中,根据局部离群因子算法自适应阈值输出故障电池。采用实车数据验证了所提方法能有效、准确地检测出故障,具有较好的可靠性与鲁棒性。

基于MLMD的电能质量扰动检测方法

针对局部均值分解(Local Mean Decomposition,LMD)算法应用于电能质量扰动检测时存在“端点效应”与滑动平均收敛速度慢,严重影响测量精度的问题,提出一种改进局部均值分解方法(Modified LMD,MLMD)。通过分段三次Hermite插值取代滑动平均法,有效改善LMD收敛慢、受平滑长度影响的弊端。为避免延拓长度不够而导致的“延拓失败”情形,在镜像延拓法的基础上结合“奇延拓”方法提出改进镜像延拓法。针对“直接法”求频率存在“毛刺现象”的弊端,文中改用希尔伯特变换(Hilbert Transform,HT)求取瞬时频率。最后,将MLMD分别应用于单一扰动信号与复合谐波信号的检测,相较传统的经验模态分解方法(Empirical Mode Decomposition,EMD),MLMD方法可有效抑制“端点效应”,同时能更准确的定位扰动信号的起止时刻,并且对高次谐波信号有更好的提取能力。