基于压缩感知成像系统的动态背景去噪

为了消除背景光强动态变化引起的非线性误差,提出了基于压缩感知成像系统的动态背景去噪算法.算法将测量中不同背景光强下所得测量值的平均值之差作为补偿系数,通过补偿系数消除动态背景噪声对压缩感知成像系统的非线性影响.仿真结果表明,在总数为900次的采样过程中,背影噪声动态变化300次时,算法能够将重建图像的峰值信噪比由29.5dB提高到62dB;在动态背景噪声的影响下,本文算法能够大大提高目标图像的可读性,提高成像质量,增加压缩感知成像系统的鲁棒性.     

基于多滤波算法融合的红外小目标检测

针对海天背景下红外视频图像序列中小目标检测的精度问题,提出一种基于多滤波融合的算法。通过分析红外小目标及背景噪声的成像特征与时空特性,将Tophat算法与改进的Robinson guard滤波器相结合,有效抑制背景、突显目标;采用自适应阈值分割提取候选目标,并使用Unger平滑滤波以及多目标关联滤波剔除噪声和伪目标。多组红外小目标图像序列的检测实验结果表明,提出的算法平均检测率高达99%,在不同场景下均有较高的检测精度。     

机车驾驶室中自适应有源降噪系统的研究

根据机车驾驶室中噪声特征的分析结果，提出了采用FIR滤波器的自适应有源降噪(ANNC)系统。为消除机车驾驶室中严重的背景噪声干扰影响，应用伪随机相关法对驾驶室中的噪声脉冲响应进行测量。通过数字仿真研究，对基于IMS算法的ANNC系统的关键参数进行了寻优。最后，在实际声场中建立了实时双通道ANNC系统并进行了相关的实验验证。仿真与实验的结果证明了本文提出的AANC系统的可行性和有效性。     

水下目标辐射噪声时间序列的非线性降噪处理

噪声是信号检测和目标识别的障碍．对水下目标辐射噪声降噪,过去常用线性的方法来对信号进行滤波。对于线性系统产生的信号,用线性滤波方法可以达到较好的效果但对非线性系统产生的信号或受非线性因素影响较大的信号,用线性滤波万法不能很好地滤除噪声,已经证明,水下目标辐射噪声信号中具有明显的非线性成分。因此,本文提出运用非线性方法对水下目标辐射噪声时间序列进行降噪处理。通过降噪前后重构吸引子、关联积分曲线斜率及信噪比的比较,表明该方法对去除水下目标辐射噪声时间序列中的噪声干扰十分有效。     

相位敏感型光时域反射传感系统光学背景噪声的产生机理及其抑制方法

相位敏感型光时域反射(Φ-OTDR)传感系统具有响应速度快、灵敏度高等优点,能够实现对微弱扰动的分布式检测,在重大设施的入侵警戒、大型工程结构的健康监测等领域具有广阔应用前景.然而,与传统的OTDR传感系统不同,Φ-OTDR系统中存在着激光器中心频率漂移、偏振相关的噪声、光纤应变与干涉强度非线性对应关系引起的测量失真等光学背景噪声,对有效信号的提取形成了不可忽视的干扰,从而限制了Φ-OTDR传感系统在实际应用环境下的传感性能.本文对这些光学背景噪声的产生机理进行了深入分析,并提出了相应的噪声抑制方法.实验结果表明,本文提出的方法可以有效抑制Φ-OTDR传感系统中的光学背景噪声,并显著提高传感系统性能.     

相干激光雷达图像散斑噪声抑制算法

提出了一种新的用于散斑噪声抑制的非线性加权均值多方向形态滤波算法。运用此算法和先前提出的散斑噪声抑制算法以及小波软阈值算法和Lee滤波算法一并对相干激光雷达图像进行了噪声抑制处理，并且对处理的结果进行了比较，证明了本算法的进步，即在不降低噪声抑制能力的前提下，较好地保持了图像边缘。     

基于Boxcar滤波器的荧光检测二氧化硫信号的去噪研究

荧光法以物质发射的荧光强度与浓度之间的线性关系进行定量分析。当采用荧光法检测大气中的二氧化硫等有害气体时,由于光电探测器等光电元件在无荧光时,仍会产生暗电流噪声,使得本底噪声信号对测量结果有直接的影响。在分析Boxcar滤波算法的基础上,运用小波滤波、EMD滤波和Boxcar滤波三种算法对淹没在本底噪声中的荧光信号进行有效提取和恢复,较之前两种滤波方法,Boxcar滤波效果更佳,对本底噪声的抑制更强,并验证了取样次数影响着荧光信号的信噪比。     

声强测量中背景噪声引起的误差

在声强测量中引起测量误差的主要原因之一是背景噪声的干扰。背景噪声引起的测量误差的大小不但与背景噪声场有关,而且与辐射体表面的吸声特性有关。本文对此问题作了理论分析。为了确定在飞机座舱内测量声强时的误差判据,对不同的材料进行了试验,给出了实用的工程判据及在强背景噪声环境下避免背景干扰的作法。     

低对比度小目标检测

对强杂波背景下的远距离目标探测,提出基于序列图像的局部自适应背景预测,获得图像背景的最佳估计。对残差图像采用能量累积及中值滤波消除背景杂波。为提高信噪比,采用带缓冲窗口的双窗滤波法使目标和背景的差别更加显著,有利于低对比度下的目标分割。最后采用改进的高阶相关方法,在不影响检测性能的情况下加快了真实目标识别的运算收敛速度,并最终实现了算法工程化,在图像局部信噪比大于0.3时,采用三阶相关时检测概率达到98%。     

红外图像掠海小目标的检测算法研究

针对红外图像掠海小目标信噪比低，且易受到水天线和背景杂波干扰的特点，提出了一种红外掠海小目标的提取与检测算法。该方法的特点是：首先采用中值滤波来减小噪声，并提出了差方和算法，用以抑制背景噪声并对目标增强；再采用了图像行扫描法有效地滤除水天线；最后通过弱化背景边缘和阈值分割等综合算法得到候选小目标。仿真结果表明，该算法达到了较好的效果。     

Pseudo-periodic surrogate test to sample time series in stochastic softening Duffing oscillator

Identification of typical noise-contaminated sample response is a hard task in a nonlinear system under stochastic background since irregularity of the sample response may come from measure noise, dynamical noise, or nonlinear effect, etc., and conventional dynamical methods are generally not useful. Here, the pseudo-periodic surrogate algorithm by Small is employed to test the sample time series in the softening Duffing oscillator under the Gaussian white noise excitation. The correlation dimensions of the noisy periodic and the noise-induced chaotic time series of the system are compared with those of their corresponding surrogate data respectively, the leading Lyapunov exponents by Rosenstein's algorithm are also presented for comparison.     

Study on the mean first-passage time and stochastic resonance of a multi-stable system with colored correlated noises

The mean first-passage time (MFPT) and the weak signal detection method of stochastic resonance (SR) on multi-stable nonlinear system under color correlated noise are studied. Using the uniform color noise approximation method, the Fokker-Planck equation of the system is obtained, and the steady-state probability density function of the multi-stable system driven by the multiplicative noise and additive noise is derived. On the basis of this, the formula of MFPT is derived, and the influence of parameters on the MFPT is analyzed. The problem of weak signal detection under color noise background is studied based on multi-stable SR. The results of simulation and experiment show that the method can effectively extract the frequency feature of weak signal in the background of color noise.     

基于随机共振进行弱信号探测的实验研究

非线性随机共振系统可利用噪声增强微弱信号检测的能力,为强噪声背景下微弱信号的检测开创了新方法.基于随机共振的基本原理设计了硬件电路系统,并将其应用于检测单频和多频微弱信号;通过输入模拟工程实际的带噪信号,采样所得的输出信号的频谱分析结果表明,利用随机共振技术可从强噪声背景下有效地提取出单频和多频弱信号.多频弱信号的有效提取拓展了基于随机共振原理的弱信号检测技术的应用领域,结合数字滤波处理技术有效地消除了低频噪声对信号识别的影响.基于随机共振的弱信号检测技术在信息识别与信息处理方面具有巨大的潜在的应用价值.     

Comparison of the Raman detection limit for IRS- and CARS-spectroscopy at excitation near molecular one-photon resonance

The nonlinear Raman methods IRS and CARS are compared according to the signal detectability at excitation of molecules under the condition of one-photon resonance. At one photon resonance it is the background contribution resulting from the scattering molecules themselves that determines the maximum attainable signal to noise ratio. These contributions acting for IRS and CARS respectively are compared. The essential difference between IRS and CARS results from a 3rd order saturation contribution to IRS, which may mask the IRS Raman signal near exact resonance while it does not contribute to CARS. This gives to CARS the preference before IRS at resonance excitation. The situation for IRS with respect to background is similar to that of spontaneous Raman scattering, where the resonance fluorescence — corresponding to the saturation contribution at IRS — masks the resonance Raman signal.     

Stochastic resonance in two kinds of asymmetric nonlinear systems with time-delayed feedback and subject to additive colored noise

This paper attempts to investigate the stochastic resonance (SR) behaviors in two kinds of asymmetric nonlinear systems with time-delayed feedback driven by additive colored noise by virtue of two-state theory, small time delay approximation, path integral approach, and unified colored-noise approximation, where asymmetric nonlinear systems include asymmetric well depth and asymmetric well width alone. The characteristics of SR in two kinds of asymmetric systems are different for different asymmetric ratios and correlated times of additive colored noise. For asymmetric well width, optimal noise intensity is independent of asymmetric ratio and correlated time, whereas for asymmetric well depth it is closely related with asymmetric ratio and correlated time. However, optimal noise intensity is closely related with feedback intensity, and time-delay for two kinds of asymmetries. Even there exists the optimal feedback intensity, time delay and correlated time to make output SNR maximum. Above clues are helpful to achieve weak signal detection under strong background noise.     

A nonlinear model for calibration of blood glucose noninvasive measurement using near infrared spectroscopy

In order to improve prediction accuracy of calibration in human blood glucose noninvasive measurement using near infrared (NIR) spectroscopy, a modified uninformative variable elimination (mUVE) method combined with kernel partial least squares (KPLS), named as mUVE–KPLS, is proposed as an alternative nonlinear modeling strategy. Under the mUVE method, high-frequency noise and matrix background can be eliminated simultaneously, which provide a optimized data for calibration in sequence; under the kernel trick, a nonlinear relationship of response variable and predictor variables is constructed, which is different with PLS that is a complex model and inappropriate to describe the underlying data structure with significant nonlinear characteristics. Two NIR spectra data of basic research experiments (simulated physiological solution samples experiment in vitro and human noninvasive measurement experiment in vivo) are introduced to evaluate the performance of the proposed method. The results indicate that, after elimination high-frequency noise and matrix background from optical absorption of water in NIR region, a high-quality spectra data is employed in calibration; and under the selection of kernel function and kernel parameter, the best prediction accuracy can be got by KPLS with Gaussian kernel compared with Spline-PLS and PLS. It is encouraging that mUVE–KPLS is a promising nonlinear calibration strategy with higher prediction accuracy for blood glucose noninvasive measurement using NIR spectroscopy.     

α稳定噪声环境下过阻尼系统中的参数诱导随机共振现象

本文采用随机模拟方法, 研究了过阻尼振子系统在α稳定噪声环境下的参数诱导随机共振现象. 结果表明, 在α噪声环境下, 调节系统参数能够诱导随机共振现象; 而且调节非线性项参数时, 随机共振效果随α稳定噪声的指数的减小而减弱, 但当调节线性项参数时, 随机共振效果则随着α稳定噪声的特征指数的减小而增强. 本文的结论在α稳定噪声环境下, 利用参数诱导随机共振原理进行弱信号检测方面具有重要的理论意义, 并有助于理解不同α稳定噪声对一般随机共振系统的共振效果的影响.     

受迫薛罗格双匣化学反应模型的随机共振

为了在理论上揭示高斯白噪声激励的薛罗格双匣化学反应模型对弱周期扰动的线性与非线性响应 ,分四态近似和两态近似两种情形 ,基于绝热近似与速率方程方法 ,解析导出线性的和非线性的敏感性以及信噪比的表达式 ,并与数值模拟结果进行比较 ,在一次谐波的意义上得到了解析结果与数值模拟结果的定量一致性 .理论上讲 ,该模型只能表现出奇次谐波的随机共振 ,但数值模拟结果也出现了二次谐波的随机共振 ,其原因可能归结为在数值模拟中有限频率的截断引入了误差 ,也可能归结为信号的高次谐波与背景噪声难以区分所致 .     

基于广义窗函数和最小二乘支持向量机的混沌背景下微弱信号检测

为了从混沌背景中检测微弱信号,研究分析了复杂非线性系统的相空间重构理论,提出了一种基于广义窗函数的最小二乘支持向量机的预测法. 该方法以广义嵌入窗为基础,利用自关联函数法确定Lorenz系统的嵌入维数和时间延迟, 实现相空间重构,结合最小二乘支持向量机建立Lorenz系统的误差预测模型, 检测微弱目标信号(瞬态和周期信号).仿真实验表明,该方法的预测模型具有较小的误差, 能够有效地从混沌背景噪声中检测出微弱目标信号,减小噪声对目标信号的影响. 与传统方法相比,在降低检测门限的同时,能够有效地提高预测的精度, 在混沌噪声下信噪比为-87.41 dB的情况下,相对于传统支持向量机方法所得的均方根误差0.049(-54.60 dB时)降低近两个数量级至0.000036123(-87.41 dB时).