基于小波变换的木材近红外光谱去噪研究

木材近红外光谱常常被一系列噪声所污染,影响光谱分析结果。为了提高近红外光谱分析精度,需要对光谱数据进行预处理。光谱导数可以消除光谱背景干扰和基线漂移等因素影响,提高光谱分辨率,但导数光谱在增强信号的同时,也使信号噪声得到增强。应用小波变换对杉木木材近红外一阶导数光谱进行去噪研究,分别采用9点平滑法、25点平滑法、非线性小波硬阈值和软阈值法、9点平滑+小波变换法和25点平滑+小波变换法对光谱数据进行去噪研究。结果显示, 小波变换能够有效去除导数光谱中的噪声信号,保留光谱中的有效信息,提高光谱信噪比,提高光谱的分析能力,在木材近红外光谱分析中具有很好的应用前景。     

Analysis of CCD moiré pattern for micro-range measurements using the wavelet transform

A new approach based on the moiré theory and wavelet transform (WT) is proposed for measuring the micro-range distance between a charge-couple-device (CCD) camera and a two-dimensional reference grating. The micro-range distance is determined by measuring the pitch of the moiré pattern image, which is digitized by a CCD camera. A one-dimensional WT algorithm is applied to estimate the pitch of the moiré pattern. Experimental results prove that this technique is very efficient and highly accurate. The moiré range finder is an economic technique for measuring a micro-range distance.     

Chemometrical strategies for feature selection and data compression applied to NIR and MIR spectra of extra virgin olive oils for cultivar identification

The possibility provided by Chemometrics to extract and combine (fusion) information contained in NIR and MIR spectra in order to discriminate monovarietal extra virgin olive oils according to olive cultivar (Casaliva, Leccino, Frantoio) has been investigated.Linear discriminant analysis (LDA) was applied as a classification technique on these multivariate and non-specific spectral data both separately and jointly (NIR and MIR data together).In order to ensure a more appropriate ratio between the number of objects (samples) and number of variables (absorbance at different wavenumbers), LDA was preceded either by feature selection or variable compression. For feature selection, the SELECT algorithm was used while a wavelet transform was applied for data compression.Correct classification rates obtained by cross-validation varied between 60% and 90% depending on the followed procedure. Most accurate results were obtained using the fused NIR and MIR data, with either feature selection or data compression.Chemometrical strategies applied to fused NIR and MIR spectra represent an effective method for classification of extra virgin olive oils on the basis of the olive cultivar.     

Tight wavelet frames for irregular multiresolution analysis

An important tool for the construction of tight wavelet frames is the Unitary Extension Principle first formulated in the Fourier-domain by Ron and Shen. We show that the time-domain analogue of this principle provides a unified approach to the construction of tight frames based on many variations of multiresolution analyses, e.g., regular refinements of bounded L-shaped domains, refinements of subdivision surfaces around irregular vertices, and nonstationary subdivision. We consider the case of nonnegative refinement coefficients and develop a fully local construction method for tight frames. Especially, in the shift-invariant setting, our construction produces the same tight frame generators as the Unitary Extension Principle.     

Multi-resolution analysis of wavelet transform on pressure fluctuations in an L-valve

A novel diagnostic method to characterize the flow patterns in an 80 mm-i.d. L-valve had been developed by using multi-resolution analysis (MRA) of wavelet transformation on the pressure fluctuation signals which were acquired from the standpipe and the horizontal part of L-valve. Parameters including the aeration rate, aeration positions, riser gas velocity and composition of binary particle mixture (194-μm and 937-μm sand particles) were used to investigate the relationship of performance of L-valve and its pressure fluctuations. By means of MRA, the original pressure fluctuations were divided into multi-scale signals. They were macro-scale, meso-scale and micro-scale successfully described the structures of gas–solid flow in the L-valve, such as the gas bubbles/slugs, dune-ripple flow, suspension particle flow, etc.     

End-point detection in potentiometric titration by continuous wavelet transform

The aim of this work was construction of the new wavelet function and verification that a continuous wavelet transform with a specially defined dedicated mother wavelet is a useful tool for precise detection of end-point in a potentiometric titration. The proposed algorithm does not require any initial information about the nature or the type of analyte and/or the shape of the titration curve. The signal imperfection, as well as random noise or spikes has no influence on the operation of the procedure.The optimization of the new algorithm was done using simulated curves and next experimental data were considered. In the case of well-shaped and noise-free titration data, the proposed method gives the same accuracy and precision as commonly used algorithms. But, in the case of noisy or badly shaped curves, the presented approach works good (relative error mainly below 2% and coefficients of variability below 5%) while traditional procedures fail. Therefore, the proposed algorithm may be useful in interpretation of the experimental data and also in automation of the typical titration analysis, specially in the case when random noise interfere with analytical signal.     

Data fusion in multivariate calibration transfer

We report the use of stacked partial least-squares regression and stacked dual-domain regression analysis with four commonly used techniques for calibration transfer to improve predictive performance from transferred multivariate calibration models. The predictive performance from three conventional calibration transfer methods, piecewise direct standardization (PDS), orthogonal signal correction (OSC) and model updating (MUP), requiring standards measured on both instruments, was significantly improved from data fusion either by stacking of wavelet scales or by stacking of spectral intervals, as demonstrated by transfer of calibrations developed on near-infrared spectra of synthetic gasoline. Stacking did not produce as significant an improvement for calibration transfer using a finite impulse response (FIR) filter, but application of SPLS regression to FIR-transferred spectra improves predictive performance of the transferred model.     

Real-time optical reconstruction of the diffused 3D object using phase information calculated by Continuous Wavelet Transform

We propose the real time optical reconstruction of a three-dimensional (3D) object from a digitally recorded hologram. Phase distribution of the recorded hologram is numerically calculated by 1-dimensional Continuous Wavelet Transform (1D-CWT) for digital reconstruction with phase only information. Also, the phase distribution of 1D-CWT transferred to the spatial light modulator (SLM) is used to obtain optically reconstructed image. It is observed that an efficient real-time analysis can be achieved, if phase of 1D-CWT is used. In the same time, optically reconstructed 3D objects obtained by only phase information are about three times brighter than bleached hologram intensity's. So the minimum power loss can be obtained. Numerical and experimental results are presented.     

超宽带聚合物调制器的优化设计

设计了一种以聚合物作为材料的低损耗、宽带宽的Mach—Zehnder光波导调制器。分析了调制器的脊波导的模式特性,设计了脊波导的结构,并使用BPM软件模拟了脊形波导的光场分布;通过对光场分布的分析,优化了脊形波导的宽度Wg,脊高6,芯层高度H。同时对聚合物调制器的电极进行了优化,包括电极宽度W和电极间距D,使得调制器有较小的导体损耗以及较好的阻抗匹配。并结合了脊波导的结构参数和电极的优化参数,给出了优化结果,它能够使微波的有效折射率与光波的有效折射率达到匹配,从而使带宽达到177GHz,导体损耗为0．2569dB／cm·GHz1／2。     

基于小波变换和高斯拟合的在线谱图综合处理方法

微小型移动式现场在线检测技术是分析仪器发展的新领域。针对复杂工作环境中谱图存在强噪声干扰、谱峰重叠、不规则峰形等严重影响仪器的定性和定量准确度的瓶颈技术,提出了一种基于小波变换和高斯拟合相结合的谱图在线综合处理方法,用自研的仪器对甲苯和全氟三丁胺两种典型化合物的谱图进行了处理,并与实验室分析仪器普遍应用的算法进行了对比分析。结果表明,综合方法能够有效解决强噪声干扰、谱峰重叠、不规则峰形问题,提高仪器的定性和定量准确性,同时能够实现数据压缩,满足仪器的在线实时检测要求。综合方法处理甲苯特征峰的平均信噪比(SNR)较移动平滑方法提高了1.3倍,峰位误差ΔM降低了3.6倍,处理全氟三丁胺谱图的数据压缩比为197∶1。