An innovation based random weighting estimation mechanism for denoising fiber optic gyro drift signal

In Interferometric Fiber Optic Gyroscope (IFOG), the diminution of random noise and drift error is a critical task. These errors degrade the performance of IFOG. In this paper, a modified adaptive Kalman gain correction (AKFG) algorithm is proposed to denoise IFOG signal. The covariance matrix of innovation sequence is estimated using weighted average window method in which the weights are randomly generated in the range [0, 1]. Innovation based random weighted estimation (IRWE)-AKFG is applied to denoise the IFOG drift signal. The Kalman gain is adaptively updated using the covariance matrix of innovation sequence. The proposed algorithm is applied for denoising IFOG signal under static and dynamic environment. Allan variance method is used to analyze and quantify the stochastic errors in IFOG sensor. The performance of the proposed algorithm is compared with Conventional Kalman filter (CKF) and the simulation results reveal that the proposed algorithm is an efficient algorithm for denoising the IFOG signal.     

Generalized total variation-based MRI Rician denoising model with spatially adaptive regularization parameters

Magnetic resonance imaging (MRI) is an outstanding medical imaging modality but the quality often suffers from noise pollution during image acquisition and transmission. The purpose of this study is to enhance image quality using feature-preserving denoising method. In current literature, most existing MRI denoising methods did not simultaneously take the global image prior and local image features into account. The denoising method proposed in this paper is implemented based on an assumption of spatially varying Rician noise map. A two-step wavelet-domain estimation method is developed to extract the noise map. Following a Bayesian modeling approach, a generalized total variation-based MRI denoising model is proposed based on global hyper-Laplacian prior and Rician noise assumption. The proposed model has the properties of backward diffusion in local normal directions and forward diffusion in local tangent directions. To further improve the denoising performance, a local variance estimator-based method is introduced to calculate the spatially adaptive regularization parameters related to local image features and spatially varying noise map. The main benefit of the proposed method is that it takes full advantage of the global MR image prior and local image features. Numerous experiments have been conducted on both synthetic and real MR data sets to compare our proposed model with some state-of-the-art denoising methods. The experimental results have demonstrated the superior performance of our proposed model in terms of quantitative and qualitative image quality evaluations.     

Effective defect detection in thin film transistor liquid crystal display images using adaptive multi-level defect detection and probability density function

A new automated inspection algorithm is proposed for detecting critical defects based on adaptive multi-level defect detection and probability density function in thin film transistor liquid crystal display (TFT-LCD) images containing a background region’s non-uniform and random noises. To improve the detecting capability for a critical-defect-detecting algorithm, the background region’s non-uniformity is eliminated using statistical values such as the mean and standard deviation of a test image. For the defect detection, the candidate defects are collected on each detection level and used to find a probability density function based on Parzen-window technique. Through simulation it was verified that the proposed method has superior capability for detecting critical defects which results in smaller brightness difference between a defect and its neighbors.     

Functional MRI using intravascular contrast agents: detrending of the relative cerebrovascular (rCBV) time course

In pharmacological fMRI experiments in animal models, blood pool contrast agents may be used to map cerebral blood volume change as a surrogate for neural activation. When the background signal drift due to contrast agent washout is non-negligible over the duration of the signal changes of interest, time-course detrending is essential for accurate interpretation of the experiment. Detrending approaches based on estimation of the background signal from a baseline period of the time course prior to pharmacological (or functional) challenge were evaluated with the aim of identifying a robust method of estimating the contrast agent washout contribution to the background signal drift. For fMRI studies in the rat, it was found that a constrained fit of a mono-exponential washout model was more accurate than a constant background approximation and unconstrained fits for experiments investigating the functional response to rapid pharmacological challenges such as cocaine and amphetamine. Moreover, the constrained fitting approach allows shorter baseline periods than unconstrained extrapolation, reducing the required duration of the experiment.     

去除乘性噪音的主成分分析算法

雷达成像系统的进一步应用依赖于对图像中噪音的有效抑制.在目前现有消除噪音方法的基础上,基于图像的局部相似性,结合主成分分析法,提出一种新的有效去除乘性噪音的滤波算法.乘性噪音经对数变换后可转化为加性噪音处理.分析了对数域中噪音的类型.首先在图像的对数域,通过非局部方法选取局部相似块作为训练样本,利用主成分分析法提取出信号的主要特征.然后基于统计理论中最小均方误差估计法给出了一种适用于图像信息的阈值原则.最后分析了变换过程引起的偏差,由对数域的偏估计得到滤波图像.数值实验验证了新算法的有效性.对比于目前提出的变分方法,新算法处理后的图像有更高的信噪比和更好的视觉效果,且具有一定的实用性.     

一种改进的多向梯度红外小目标检测方法

红外小目标检测技术是红外搜索与跟踪系统（IRST）、红外警戒等系统中的一项关键技术。常规小目标检测算法存在对椒盐噪声、高亮背景边缘敏感等不足之处,针对上述问题,对天空背景红外弱小目标的几何和灰度特性进行了深入研究,在此基础上提出了基于多向梯度的红外小目标检测的新方法,采用了一种非对称的梯度算子,重新定义了梯度检测函数,提高了对小目标的响应。仿真实验结果表明,该方法能够较好地抑制高亮云层边缘和椒盐噪声,保证了较高的检测率和较低的虚警率,又便于在FPGA上实现,满足系统对实时性的要求。     

基于独立成分分析的高光谱图像异常检测

针对高光谱图像中背景及目标先验知识未知条件下的异常目标检测问题,提出了一种基于独立成分分析(ICA)的异常探测算法.首先估计原始数据的虚拟维(VD)以确定要分离的独立成分个数,在此基础上进行快速独立成分分析(FastICA),然后基于平均局部奇异度选择含异常信息较多的独立成分,最后使用丰度量化算法得到异常目标的丰度图像...     

葡萄糖注射液中不溶性微粒的识别算法研究

针对传统人工目测法以及光阻检测法在葡萄糖注射液中不溶性微粒识别判断效率较低现象,提出了一种背景差预处理与改进的Mean Shift算法相结合的微粒识别算法。利用高分辨率数码相机采集在一定照度下葡萄糖注射液中不溶性微粒产生散射光的序列数字图像。对序列数字图像进行预处理及并标定数字图像有效检测区域。建立序列数字图像背景模型并对背景噪声与随机噪声进行抑制处理。采用改进的Mean Shift算法对葡萄糖注射液中不溶性微粒进行目标跟踪。通过对序列数字图像中被检测对象的分割与特征提取,实现了葡萄糖注射液中粒径在25μm以上不溶性微粒的识别。     

Damage detection in composite panels based on mode-converted Lamb waves sensed using 3D laser scanning vibrometer

This paper introduces damage identification approach based on guided ultrasonic waves and 3D laser Doppler vibrometry. The method is based on the fact that the symmetric and antisymmetric Lamb wave modes differ in amplitude of the in-plane and out-of-plane vibrations. Moreover, the modes differ also in group velocities and normally they are well separated in time. For a given time window both modes can occur simultaneously only close to the wave source or to a defect that leads to mode conversion. By making the comparison between the in-plane and out-of-plane wave vector components the detection of mode conversion is possible, allowing for superior and reliable damage detection. Experimental verification of the proposed damage identification procedure is performed on fuel tank elements of Reusable Launch Vehicles designed for space exploration. Lamb waves are excited using low-profile, surface-bonded piezoceramic transducers and 3D scanning laser Doppler vibrometer is used to characterize the Lamb wave propagation field. The paper presents theoretical background of the proposed damage identification technique as well as experimental arrangements and results.     

浅海小目标前向散射的非互易时反检测方法

将无源非互易时反(NRTR)用于浅海小目标前向散射探测海上实验研究,提出了一种基于频域加权松弛的信道估计算法(WRELAX)和时反方法相结合的小目标探测算法(WRELAX-NRTR)。该算法采用高精度的信道响应估计实现信道的空-时匹配,利用目标引起的时反声场结构的畸变检测前向散射信号对声场的扰动实现目标探测,并通过阈值法实现目标判决。在15~30 kHz的信号频率范围内进行海上实验,数据的处理结果表明,提出的WRELAX-NRTR探测算法与传统的NRTR算法相比,环境背景的扰动幅度从10 dB减小到4 dB,从而降低了虚警率,另外该算法对海洋环境噪声、目标聚焦深度、目标穿越距离和信道动态变化等均具有良好的鲁棒性。      

Automatic detection of bioacoustics impulses based on kurtosis under weak signal to noise ratio

Passive acoustic monitoring (PAM) of marine mammal vocalizations has been efficiently used in a wide set of applications ranging from marine wildlife surveys to risk mitigation of military sonar emissions. The primary use of PAM is for detecting bioemissions, a good proportion of which are impulse sounds or clicks. A click detection algorithm based on kurtosis estimation is proposed as a general automatic click detector. The detector works under the assumption that click trains are embedded in stochastic but Gaussian noise. Under this assumption, kurtosis is used as a statistical test for detection. The algorithm explores acoustic sequences with the optimal frequency bandwidth for focusing on impulse sounds. The detector is successfully applied to field observations, and operates under weak signal to noise ratios and in presence of stochastic background noise. The algorithm adapts to varying click center frequency. Kurtosis appears as a promising approach to detect click trains, alone or in combination with other clicks detector, and to isolate individual clicks.     

An embedded pilot power based channel estimation and low-complexity feedback equalization scheme for OTFS system

Orthogonal Time Frequency Space modulation (OTFS) has evolved as an astounding modulation technique for high-speed communication in a doubly dispersive channel. In any wireless communication system, channel estimation and equalization are essential at the receiver to recover the transmitted data. To accomplish this for the emerging OTFS based systems, a modified embedded pilot-based channel estimation technique and low complexity feedback equalization algorithm for integer Doppler shifts in the delay-Doppler domain are proposed in this paper. Our channel estimation scheme exploits embedded-pilot arrangement, and the symbol equalization relies on the Interference calculation and its mitigation iteratively. To achieve this we contemplate a prudent arrangement of symbols in the OTFS frame in such a way that the Guard symbols prevent the interference between data symbols and the pilot symbol at the receiver. Two distinct lumps of received data of the same OTFS frame will be engaged in channel estimation and data detection. An analytical expression of the theoretical Cramer Rao Lower Bound (CRLB) is derived and plotted for the proposed channel estimation scheme. The attained simulation results for Bit-Error-Rate (BER) under the proposed scheme show a significant error rate improvement over the Minimum Mean Squared Error (MMSE) equalization algorithm. Further, a lower computational complexity is also achieved in comparison with modified MMSE detection and MP detection algorithms.     

混沌噪声背景下微弱脉冲信号的检测及恢复

构建了一种在混沌噪声背景下检测并恢复微弱脉冲信号的模型.首先,基于混沌信号的短期可预测性及其对微小扰动的敏感性,对观测信号进行相空间重构、建立局域线性自回归模型进行单步预测,得到预测误差,并利用假设检验方法从预测误差中检测观测信号中是否含有微弱脉冲信号.然后,对微弱脉冲信号建立单点跳跃模型,并融合局域线性自回归模型,构成双局域线性(DLL)模型,以极小化DLL模型的均方预测误差为目标进行优化,采用向后拟合算法估计模型的参数,并最终恢复出混沌噪声背景下的微弱脉冲信号.仿真实验结果表明本文所建的模型能够有效地检测并恢复出混沌噪声背景中的微弱脉冲信号.     

电力设备红外图像的小波阈值去噪方法研究

对于电力设备的红外图像自动检测系统,图像去噪是非常关键的。针对传统的小波硬、软阈值函数在去噪时存在的不足,在传统小波硬、软阈值函数的基础上对其进行了改进。改进的阈值函数克服了硬阈值函数不连续的缺陷,改善了软阈值函数具有恒定偏差的不足,并引入了两个变量,具有一定的灵活性。同时还使用了一种新的分层阈值选择函数代替统一阈值方法,以改善实际应用的效果。实验结果表明:改进的小波阈值去噪方法在视觉效果、峰值信噪比和均方误差方面都优于传统的硬、软阈值去噪方法;改进的小波阈值去噪方法可以运用到红外图像自动检测系统中,使系统具有更好的去噪效果。     

Autocorrelation based denoising of manatee vocalizations using the undecimated discrete wavelet transform

Recent interest in the West Indian manatee (Trichechus manatus latirostris) vocalizations has been primarily induced by an effort to reduce manatee mortality rates due to watercraft collisions. A warning system based on passive acoustic detection of manatee vocalizations is desired. The success and feasibility of such a system depends on effective denoising of the vocalizations in the presence of high levels of background noise. In the last decade, simple and effective wavelet domain nonlinear denoising methods have emerged as an alternative to linear estimation methods. However, the denoising performances of these methods degrades considerably with decreasing signal-to-noise ratio (SNR) and therefore are not suited for denoising manatee vocalizations in which the typical SNR is below 0 dB. Manatee vocalizations possess a strong harmonic content and a slow decaying autocorrelation function. In this paper, an efficient denoising scheme that exploits both the autocorrelation function of manatee vocalizations and effectiveness of the nonlinear wavelet transform based denoising algorithms is introduced. The suggested wavelet-based denoising algorithm is shown to outperform linear filtering methods, extending the detection range of vocalizations.     

An improved TOA estimation algorithm based on denoised MVDR for B5G positioning

This paper proposes an improved minimum variance distortionless response (MVDR) based TOA estimation algorithm for 5G NR signals under multipath environments. The proposed algorithm achieves high resolution by exploiting a large number of subcarriers of 5G signals and reduces the dimension of the covariance matrix involved in MVDR substantially by utilizing a novel smoothing scheme. Since MVDR requires a relatively high signal-to-noise ratio (SNR), a denoising method is used to improve the TOA estimation performance. Simulation results show that the proposed algorithm achieves much higher resolution than the Bartlett beamformer (BF) and the TOA estimation accuracy remains high over a wide range of SNRs.     

一种用于谐波检测过程中消除谐波背景方法的研究

针对可调谐半导体激光吸收光谱（TDLAS）连续检测技术中,二次谐波背景信号存在漂移的现象,提出改变激光器中心电流实时提取背景信号,以消除连续检测过程中背景信号的漂移对浓度反演的影响。依据波长调制理论推导了二次谐波背景信号的理论表达式,并分析了实际情况下影响二次谐波背景信号的因素。给出激光器在不同工作温度时电流和输出光强度之间的关系曲线,并分析了改变激光器中心电流实时提取背景信号的可行性。结合背景信号搜索方法设计了基于LabVIEW的背景信号提取流程图。设计以氨气为检测对象的TDLAS实验系统,选取了氨气的吸收谱线以及对应的吸收中心电流。在激光器电流全工作区间内只存在唯一吸收峰的情况下,确定实验中各参数的数值及搜索背景的电流范围。实验结果表明：该方法可实时提取谐波背景信号。结合线性最小二乘法拟合反演可有效地减小检测误差及背景信号对浓度反演精度的影响,提高浓度的检测精度及准确性。在连续检测实验中,反演浓度的标准差由2.688 3降到1.856 1,减小背景信号漂移对检测浓度准确性的影响,提高了连续检测的准确性。     

Relevant modes selection method based on Spearman correlation coefficient for laser signal denoising using empirical mode decomposition

Empirical mode decomposition (EMD) is a recently proposed nonlinear and nonstationary laser signal denoising method. A noisy signal is broken down using EMD into oscillatory components that are called intrinsic mode functions (IMFs). Thresholding-based denoising and correlation-based partial reconstruction of IMFs are the two main research directions for EMD-based denoising. Similar to other decomposition-based denoising approaches, EMD-based denoising methods require a reliable threshold to determine which IMFs are noise components and which IMFs are noise-free components. In this work, we propose a new approach in which each IMF is first denoised using EMD interval thresholding (EMD-IT), and then a robust thresholding process based on Spearman correlation coefficient is used for relevant modes selection. The proposed method tackles the problem using a thresholding-based denoising approach coupled with partial reconstruction of the relevant IMFs. Other traditional denoising methods, including correlation-based EMD partial reconstruction (EMD-Correlation), discrete Fourier transform and wavelet-based methods, are investigated to provide a comparison with the proposed technique. Simulation and test results demonstrate the superior performance of the proposed method when compared with the other methods.     

基于统计方法的动态光谱差值提取

为实现血液成分无创检测,针对动态光谱(dynamic spectrum, DS)时域单拍提取法存在的不足,提出时域差值提取法,即顺序提取各波长对数光电脉搏波上相距理想差值间隔的两对应采样点的绝对差值构成差值DS,利用统计方法优选若干有效差值DS进行叠加平均获取最终DS输出。对48例志愿者的实验数据分别利用差值提取法和单拍提取法进行DS提取,结果表明：差值提取法提取DS的去噪效果略优于单拍提取法;对单一个体可获取有效DS的平均个数由48个改善为130个,有效DS间均方误差的平均值由0.39改善为0.006,运算速度提升了近20倍。新方法显著提高了DS提取的质量。