小波变换在光谱特征提取方面的应用

人们在处理高光谱图像时一般要对一些典型地物进行光谱分析、特征波段的提取,以便提取出最大量的有效信息,剔除无用或冗余的信息,然后再进行分类识别.采用小波变换的分析方法,选用合适的小波进行分解,根据分解后的高频分量中包含的重要信息,利用局部相邻的正负极值点找出对应于原始光谱曲线上每个吸收带的左右边界;利用局部过零点,即可比较精确的提取出各个吸收带的中心波长.该方法比传统的光谱特征提取方法更简洁、有效,实验证明为一种比较理想的光谱特征提取方法.     

基于小波特征的星系光谱分类

提出了一种新的星系光谱分类方法.首先,对原始光谱进行四级小波分解,选择主要包含谱线信息的第四级小波系数作为光谱的小波特征;然后,利用主分量分析对光谱的小波特征进行特征压缩,得到光谱的识别特征;最后,利用Fisher线性判别分析实现分类.该方法能够在红移值未知的情况下,对流量未定标的星系光谱进行识别.通过实验与其他几种分类方法进行了比较.实验结果表明,本文方法具有较强的鲁棒性,在流量未定标情况下的识别效果优于其他几种分类方法.     

一种基于小波变换的YOLOv5火灾检测改进算法

由于复杂环境下类烟火物体的干扰,常导致火灾检测误判。为了提高图像中火灾信号的检测精度,减少火灾误报,利用传统光谱分析在火灾图像检测技术中的优势,提出了一种基于小波变换的YOLOv5火灾检测改进算法。该算法利用二维Haar小波变换提取图像的光谱特征,将其输入到YOLOv5s的主干网络CSPDarknet中,与卷积层进行通道上的特征融合,增强烟火的纹理细节特征;通过嵌入CA注意力机制的CAC3模块,对融合小波特征后的网络层的位置信息进行增强,提高网络的信息提取和定位能力;为明确衡量边界框宽高的真实差,平衡烟火难易样本,采用α-EIOU损失函数替换原本的CIOU,提高框定位准确性。在公开的火灾数据基础上结合自制火灾数据构建火灾数据集,并进行模型训练和推理。实验结果表明,改进后算法的mAP比原YOLOv5s提升了2.3%,实现了对火灾场景烟火目标较好的检测效果。     

小波分析在煤矿胶带机滚动轴承故障诊断中的应用

本文介绍了小波变换理论及小波分析的特点,对小波变换应用于煤矿滚动轴承故障诊断的方法和滚动轴承故障特征频率的计算方法进行了研究。根据煤矿胶带机滚动轴承的振动特征,应用小波分解重构的方法对煤矿胶带机滚动轴承的现场故障振动信号进行分析,验证小波变换在提取滚动轴承早期故障信号特征方面的有效性。     

一种基于小波神经网络的缺陷检测新算法

运用多分辨分析对缺陷检测信号进行小波分解和重构,消除信号中的噪声,再基于多分辨分析提取检测信号的特征量。选取B-小波为小波基构建了断丝定量识别的小波神经网络模型,并运用该小波神经网络对缺陷检测信号进行了检测,试验结果证明了小波神经网络能够较为理想地完成缺陷信号的智能化检测,对断丝位置及断丝根数能实现无人工干预的自动识别,比传统的断丝识别方法准确率提高了很多。     

M带小波变换在图像压缩中的应用

利用小波变换进行图像压缩一直是人们感兴趣的问题。本文根据M带小波变换理论，分析了图像经M带小波变换后的数据结构特性，利用人眼视觉特征，提出了图像的压缩编码算法。实验结果证实，该方法对图像压缩行之有效。     

基于小波包分解的图像信息隐写盲检测

基于小波包分解,提出了一类新的具有较高检测正确率的图像信息隐写盲检测方法.首先对图像进行小波包分解得到多个子带,从子带系数以及图像像素中提取直方图特征函数多阶绝对矩作为特征,然后对提取的特征进行预处理并设计BP神经网络分类器进行分类.针对LSB、SS、Jsteg、F5及MB等典型隐写的实验表明:此方法相比现有的典型盲检测方法,正确检测率提高约7.5%～17.2%,且具有更好的通用性.此外,还讨论了整数和非整数小波包分解对检测结果的影响.     

多光谱图像融合中小波分解层数研究

首先根据小波变换原理,采用Db9小波基函数,对多组多光谱图像分别进行1～5层小波分解,然后根据小波逆变换原理对系数融合后的图像进行逆变换,得到不同小波分解层的融合图像.最后,利用图像质量评价方法信息熵、标准差、互信息、以及图像融合质量综合评价方法等,对不同分解层下的融合图像进行了评价.实验结果表明:小波变换在一层小波分解时,图像融合效果最佳.     

利用引入瞳孔位置的Gabor小波网络进行人脸识别

Gaber小波网络能很好地提取图像特征和进行图像表达,本文提出将瞳孔位置信息引入到Gaber小波网络的人脸特征提取中以提高提取效率.该瞳孔位置信息用于两个方面,一是在网络优化时利用瞳孔位置构造T形的小波初始位置分布,使得在小波数目一定的情况下识别信息的提取更高效;二是在小波网络的参数再确定时由瞳孔位置提供定位信息从而大大简化求参步骤.本文采用Gabor小波网络提取出人脸特征后再用核联想记忆法进行分类.实验结果表明,瞳孔位置的利用提高了人脸特征提取的效率;此外,与欧氏距离、归一化互相关和最近特征线(NFL)这些方法相比,核联想记忆法具有更好的识别率.     

基于小波变换的噪声及背景同时去除方法在血糖近红外无创检测中的应用

对于血糖近红外无创检测.光谱信号中的各种噪声以及水分等物质的强吸收产生的背景信号,影响了光谱定量校正模型的预测精度.利用小波变换,可将光谱信号分解为多尺度的近似成分与细节成分,根据无用信息变量消除判据可判定代表背景信息的高尺度近似成分及代表噪声的低尺度细节系数,去除相应的成分即可同时去除光谱信号中的背景与噪声.本文将这种小波变换与无用信息变量消除判据相结合的预处理方法应用干人体血糖无创检测研究中,并分析了该方法对不确定因素较多的复杂光谱模型的适用性问题.实验结果表明应用小波变换结合无用信息变量消除判据的方法可以有效地同时去除血糖无创检测近红外光谱信号中的背景信息和噪声,提高光谱定量校正模型的预测精度,对于人体血糖无创检测具有重要应用价值.     

Orthogonal complex filter banks and wavelets: some properties anddesign

Previous wavelet research has primarily focused on real-valued wavelet bases. However, complex wavelet bases offer a number of potential advantageous properties. For example, it has been suggested that the complex Daubechies wavelet can be made symmetric. However, these papers always imply that if the complex basis has a symmetry property, then it must exhibit linear phase as well. In this paper, we prove that a linear-phase complex orthogonal wavelet does not exist. We study the implications of symmetry and linear phase for both complex and real-valued orthogonal wavelet bases. As a byproduct, we propose a method to obtain a complex orthogonal wavelet basis having the symmetry property and approximately linear phase. The numerical analysis of the phase response of various complex and real Daubechies wavelets is given. Both real and complex-symmetric orthogonal wavelet can only have symmetric amplitude spectra. It is often desired to have asymmetric amplitude spectra for processing general complex signals. Therefore, we propose a method to design general complex orthogonal perfect reconstruct filter banks (PRFBs) by a parameterization scheme. Design examples are given. It is shown that the amplitude spectra of the general complex conjugate quadrature filters (CQFs) can be asymmetric with respect the zero frequency. This method can be used to choose optimal complex orthogonal wavelet basis for processing complex signals such as in radar and sonar     

一种新的小波消噪阈值的估计方法

小波多尺度分解是一种有效的信号去噪方法，对于非平稳信号的消噪，主要是选取合适的小波及每层小波系数的阈值。基于传统的阈值去噪方法，在分析研究了它们的优缺点之后，本着改进滤波效果，提高去噪质量的目的，提出了一种改进方案。该改进方案克服了传统阈值去噪方法的缺陷，并适用于进一步的自适应滤波的需求，仿真试验证实了该改进方案的有效性和优越性。     

一种新的小波图像去噪方法

小波图像去噪已经成为目前图像去噪的主要方法之一，目前的研究主要集中于如何选取阈值使去噪达到较好的效果。边缘信息是图像最为有用的高频信息，在图像去噪的同时，应尽量保留图像的边缘信息，基于这一思想，提出一种新的小波图像去噪方法。用数学形态学算子对图像小波变换后的小波系数进行处理，以去除具有较小支持域的噪声，保留具有连续支持域的边缘。实验结果表明，与普通的小波阈值去噪方法相比，该方法不但可以保留图像的边缘信息，而且能提高去噪后图像的峰值信噪比2～5dB，提高信噪比6～10dB。     

基于小波变换的空间目标图像去噪方法

通过对空间目标图像的特性进行分析,提出一种针对星空背景图像在保留恒星同时去除混合噪声的方法.该方法首先利用小波局部模极大值的多尺度相关性检测出图像边缘,再利用基于梯度分析的改进阈值方法对非边缘小波系数进行萎缩,最后由小波系数重构去噪后图像.实验证明该方法能够有效地去除高斯和椒盐混合噪声,使图像峰值信噪比提高5-10dB,并较好地保留图像边缘和有效恒星信息.     

基于小波变换的苹果汁多光程近红外光谱信息提取研究

利用浸入式光纤采集鲜榨苹果汁分别在5mm、10mm、15mm和20mm光程下的透/反射近红外光谱,实现对苹果汁中糖度(可溶性固形物,SSC)和酸度(pH值)的定量预测.结果表明,SSC和pH具有不同的最佳光程长,分别为5mm和20mm.为了兼顾各待测量对象的浓度范围和各组分的最佳光程长,从而提高模型的性能,采用多光程光谱混合建模,研究了多光程光谱信息的提取方法.采用原始光谱直接展开所建的模型虽然能有效利用多光程光谱的信息,但增加了模型复杂度,致使建模时间增长.因此,提出了两种基于小波变换的信息提取方法,它们在高效提取多光程信息的同时,能显著缩短建模时间并简化模型.其中基于展开光谱的小波近似系数建立的模型性能最优,SSC和pH值模型的SECV值分别达到0.4761oBrix和0.0779.     

Fuzzy wavelet packet based feature extraction method and its application to biomedical signal classification

In this paper, we develop an efficient fuzzy wavelet packet (WP) based feature extraction method for the classification of high-dimensional biomedical data such as magnetic resonance spectra. The key design phases involve: 1) a WP transformation mapping the original signals to many WP feature spaces and finding optimal WP decomposition for signal classification; 2) feature extraction based on the optimal WP decomposition; and 3) signal classification realized by a linear classifier. In contrast to the standard method of feature extraction used in WPs, guided by the criteria of signal compression or signal energy, our method is used to extract discriminatory features from the WP coefficients of the optimal decomposition. The extraction algorithm constructs fuzzy sets of features (via fuzzy clustering) to assess their discriminatory effectiveness. This paper includes a number of numerical experiments using magnetic resonance spectra. Classification results are compared with those obtained from common feature extraction methods in the WP domain.     

Wavelet transform threshold noise reduction methods in the oil pipeline leakage monitoring and positioning system

This letter investigates the wavelet transform, as well as the principle and the method of the noise reduction based on wavelet transform, it chooses the threshold noise reduction, and discusses in detail the principles, features and design steps of the threshold method. Rigrsure, heursure, sqtwolog and minimization four kinds of threshold selection method are compared qualitatively, and quantitatively. The wavelet analysis toolbox of MATLAB helps to realize the computer simulation of the signal noise reduction. The graphics and calculated standard deviation of the various threshold noise reductions show that, when dealing with the actual pressure signal of the oil pipeline leakage, sqtwolog threshold selection method can effectively remove the noise. Aiming to the pressure signal of the oil pipeline leakage, the best choice is the wavelet threshold noise reduction with sqtwolog threshold. The leakage point is close to the actual position, with the relative error of less than 1%.     

S-transform-based intelligent system for classification of power quality disturbance signals

In this paper, a new approach is presented for the detection and classification of nonstationary signals in power networks by combining the S-transform and neural networks. The S-transform provides frequency-dependent resolution that simultaneously localizes the real and imaginary spectra. The S-transform is similar to the wavelet transform but with a phase correction. This property is used to obtain useful features of the nonstationary signals that make the pattern recognition much simpler in comparison to the wavelet multiresolution analysis. Two neural network configurations are trained with features from the S-transform for recognizing the waveform class. The classification accuracy for a variety of power network disturbance signals for both types of neural networks is shown and is found to be a significant improvement over multiresolution wavelet analysis with multiple neural networks.     

一种改进的小波基线校正方法的研究

通过对小波变换基线校正中最佳分解尺度方法的研究,提出了一种新的基于小波变换的最佳分解尺度确定方法,不但有效地提高了基线校正效果,而且具有简单、快速的优点.将该方法应用于血糖光谱数据预处理中进行基线校正,取得了较好的效果.通过人体口服葡萄糖耐量试验(OGTT)得到人体无创检测近红外光谱和对应血糖浓度值,然后采用该方法对上述光谱进行基线校正并建立多元回归模型,采用交互验证的方式对模型及基线校正的效果进行了评价.实验结果表明,血糖浓度预测值和参考值间的相关系数为0.75,预测均方根误差(RMSEP)为1.36 mmol/L,与原始光谱预测结果和其他小波分解尺度下的预测结果相比,RMSEP降低了将近39%,相关系数提高了0.64,预测精度得到较大幅度提高.     

Using wavelet transform and fuzzy neural network for VPC detection from the Holter ECG

A novel method for detecting ventricular premature contraction (VPC) from the Holter system is proposed using wavelet transform (WT) and fuzzy neural network (FNN). The basic ideal and major advantage of this method is to reuse information that is used during QRS detection, a necessary step for most ECG classification algorithm, for VPC detection. To reduce the influence of different artifacts, the filter bank property of quadratic spline WT is explored. The QRS duration in scale three and the area under the QRS complex in scale four are selected as the characteristic features. It is found that the R wave amplitude has a marked influence on the computation of proposed characteristic features. Thus, it is necessary to normalize these features. This normalization process can reduce the effect of alternating R wave amplitude and achieve reliable VPC detection. After normalization and excluding the left bundle branch block beats, the accuracies for VPC classification using FNN is 99.79%. Features that are extracted using quadratic spline wavelet were used successfully by previous investigators for QRS detection. In this study, using the same wavelet, it is demonstrated that the proposed feature extraction method from different WT scales can effectively eliminate the influence of high and low-frequency noise and achieve reliable VPC classification. The two primary advantages of using same wavelet for QRS detection and VPC classification are less computation and less complexity during actual implementation.