Defect detection in textile fabric images using subband domain subspace analysis

In this work, a new model that combines the concepts of wavelet transformation and subspace analysis tools, like independent component analysis (ICA), topographic independent component analysis (TICA), and Independent Subspace Analysis (ISA), is developed for the purpose of defect detection in textile images. In previous works, it has been shown that reduction of the textural components of the textile image by preprocessing has increased the performance of the system. Based on this observation, in the present work, the aforementioned subspace analysis tools are applied to subband images. The feature vector of a subwindow of a test image is compared with that of a defect-free image in order to make a decision. This decision is based on a Euclidean distance classifier. The increase performance that results from using wavelet transformation prior to subspace analysis has been discussed in detail. While it has been found that all subspace analysis methods lead to the same detection performances, as a further step, independent subspace analysis is used to classify the detected defects according to their directionalities.     

Bayesian texture classification and retrieval based on multiscale feature vector

This paper proposes a supervised multiscale Bayesian texture classifier. The classifier exploits the dual-tree complex wavelet transform (DT-CWT) to obtain complex-valued multiscale representations of training texture samples for each texture class. The high-pass subbands of DT-CWT decomposition of a texture image are used to form a multiscale feature vector representing magnitude and phase features. For computational efficiency, the dimensionality of feature vectors is reduced using principal component analysis (PCA). The class conditional probability density function of low-dimensional feature vectors for each texture class is then estimated by using Parzen-window estimate with identical Gaussian kernels and is used to represent the texture class. A query texture image is classified as the corresponding texture class with the highest a posteriori probability according to a Bayesian inferencing. The superior performance and robustness of the proposed classifier is demonstrated for classifying texture images from image databases. The proposed multiscale texture feature vector extracted from both magnitude and phase of DT-CWT subbands of a query image is also shown to be effective for texture retrieval.     

Improved support vector classification using PCA and ICA feature space modification

An approach that unifies subspace feature selection and optimal classification is presented. Independent component analysis (ICA) and principal component analysis (PCA) provide a maximally variant or statistically independent basis for pattern recognition. A support vector classifier (SVC) provides information about the significance of each feature vector. The feature vectors and the principal and independent component bases are modified to obtain classification results which provide lower classification error and better generalization than can be obtained by the SVC on the raw data and its PCA or ICA subspace representation. The performance of the approach is demonstrated with artificial data sets and an example of face recognition from an image database.     

基于独立成分分析的表面缺陷特征提取与识别方法

为了提取表面缺陷图像特征,常对图像进行线性变换,但通常的wavelet变换、Gabor变换及其基函数都是预先定义和不变的,不能适应于缺陷图像的特点.为此提出基于独立成分分析(ICA)和拓扑独立成分分析(TICA)的特征提取方法,并将其应用于冷轧带钢表面缺陷自动识别.首先利用ICA和TICA从缺陷集中自适应地估计出基函数和滤波器,这些基适应于缺陷图像的特点;然后用与基对应的滤波器对缺陷图像滤波,提取滤波响应作为特征向量;最后用支持向量机对样本进行分类识别.该方法建立在对缺陷集无监督学习的基础上,能够自适应地提取缺陷图像的显著特征,且计算简单,可并行处理.实验结果表明,文中方法对形状类缺陷、纹理类缺陷及其他缺陷的识别率都非常高,总体识别率可达95.52％.     

一种模拟电路的支持向量机故障诊断方法

系统地提出了模拟电路的最小二乘小波支持向量机故障诊断方法。从测试点得到各种故障状态下的输出电压信号，对输出电压信号进行小波去噪，对信号进行小波分解获取多尺度的低频系数和高频系数，并对小波系数进行处理从而提取出故障特征量，以此作为学习样本来训练最小二乘小波支持向量机，确定其模拟电路故障诊断的模型。雷达系统电路仿真结果表明了模拟电路的小波变换和最小二乘小波支持向量机故障诊断方法取得了较好的效果。     

基于LDA和小波分解的肺音特征提取方法

针对用小波分解提取肺音特征后特征向量维数较高的问题,提出了一种结合线性判别分析和小波分解的肺音特征提取方法。在该方法中,首先对肺音信号进行小波分解,然后将小波分解得到的小波系数转化成小波能量特征向量,接着使用线性判别分析法对该特征向量进行降维处理,得到新的低维特征向量,最后用SVM对低维特征进行识别。在实验中,选取了三种肺音信号：正常肺音、爆裂音、哮鸣音,用所提出的方法将8维的小波能量特征降为2维特征,在2维特征上进行了分类识别,并和降维之前的结果进行了比较,实验结果表明利用线性判别分析对小波能量特征降维后极大地提高了识别精度。同时,和其他几种典型的肺音特征提取方法进行了比较,实验结果表明结合线性判别分析与小波分解的特征提取方法得到了更高的识别精度。     

基于三次B样条小波和2DFFT-2DLDA的人脸识别

提出了一种基于三次B样条小波和2DFFT-2DLDA的人脸识别方法,用三次B样条小波对人脸图像进行多层分解,得到一幅低频子图和3幅边缘细节子图,选取其中两幅效果最好的子图进行二维傅里叶变换后将其连接形成一个特征向量,然后进行2DLDA处理产生最终的特征表达,最后使用最近邻法进行分类。在JAFFE和Yale人脸库中的实验表明算法具有比频谱脸算法和Gabor-2DLDA算法更高的识别率,同时具有很低的算法复杂度。     

基于多分辨Fourier-Mellin变换的剪纸纹样识别算法

提出一种基于多分辨Fourier-Mellin的剪纸纹样识别算法。该算法先对剪纸纹样图像进行Fourier-Mellin变换,再对变换后的图像通过小波变换计算出各层方差和均值,得到剪纸纹样不同子带的特征值,应用支持向量机对剪纸纹样进行识别。实验证明,该方法不仅具有平移、旋转和尺度不变性,而且适用于有夸张艺术变形的剪纸纹样识别。     

基于ICA和SVM的SAR图像特征提取与目标识别

提出一种利用独立分量分析和支持向量机的合成孔径雷达图像特征提取与目标识别方法。对图像小波分解后提取低频子带图像,对低频子带图像进行独立分量分析提取特征向量,利用支持向量机对特征向量分类完成目标识别。将该方法用于MSTAR数据中的3类目标识别,识别率最高可达96.92%。实验结果表明,该方法是一种有效的合成孔径雷达图像特征提取与目标识别方法。     

HIRBIR: A hierarchical approach to region-based image retrieval

This paper proposes a hierarchical approach to region-based image retrieval (HIRBIR) based on wavelet transform whose decomposition property is similar to human visual processing. First, automated image segmentation is performed fast in the low-low (LL) frequency subband of the wavelet domain that shows the desirable low image resolution. In the proposed system, boundaries between segmented regions are deleted to improve the robustness of region-based image retrieval against segmentation-related uncertainty. Second, a region feature vector is hierarchically represented by information in all wavelet subbands, and each feature component of a feature vector is a unified color–texture feature. Such a feature vector captures well the distinctive features (e.g., semantic texture) inside one region. Finally, employing a hierarchical feature vector, the weighted distance function for region matching is tuned meaningfully and easily, and a progressive stepwise indexing mechanism with relevance feedback is performed naturally and effectively in our system. Through experimental results and comparison with other methods, the proposed HIRBIR shows a good tradeoff between retrieval effectiveness and efficiency as well as easy implementation for region-based image retrieval.     

应用复小波和独立成分分析的人脸识别

结合双树复小波变换（DT-CWT）和独立成分分析（ICA）提出了一种人脸识别新方法。该方法首先应用双树复小波变换提取图像的特征向量,接着通过主成分分析（PCA）降低特征向量的维数,在此基础上应用独立成分分析提取统计上独立的特征向量,然后基于相关系数的分类器对特征向量进行分类。双树复小波变换具有方向与尺度选择性,并能有效的保持图像的频域信息,其与独立成分分析相结合提取的特征具有良好的分类性能。在ORL和AR人脸图像数据库上进行算法验证的结果表明该方法的有效性。     

非线性模拟电路故障诊断的小波领袖多重分形分析方法

针对非线性模拟电路故障的复杂性和非线性,提出一种基于小波领袖多重分形分析和支持向量机的故障诊断方法.首先,采用小波领袖方法对从测试节点采集的信号进行多重分形分析,并将提取的多重分形谱特征构成特征集;然后将特征集输入支持向量机,利用支持向量机的分类功能对电路的模式状态作出决策;最后,视频放大器电路故障诊断实验验证了该方法的有效性.     

基于小波包分析和SVM的直肠感知功能预测模型研究

根据肠道蠕动机制提出一种应用于反馈式人工肛门括约肌的直肠感知功能预测模型.该模型通过小波包分析,将结肠收缩压力信号的能量分层作为特征向量,采用支持向量机进行模式识别.仿真实验中,首先直接将收缩压力信号幅值作为特征向量,分别采用BP网络、支持向量机进行模式识别,随后将所提出的预测模型与两种方法进行比较.仿真结果说明,基于...     

Log-Gabor小波和分数阶多项式KPCA的火焰图像状态识别

目的 为了进一步提高锅炉燃烧火焰图像状态识别的性能,提出了一种基于Log-Gabor小波和分数阶多项式核主成分分析(KPCA)的火焰图像状态识别方法。方法 首先利用Log-Gabor滤波器组对火焰图像进行滤波,提取滤波后图像的均值和标准差,并构成纹理特征向量。然后使用分数阶KPCA方法对纹理特征向量进行降维,并将降维后的纹理特征向量输入支持向量机进行分类。结果 本文与基于Log-Gabor小波特征提取以及2种基于Gabor小波特征提取的方法相比,本文方法的分类识别正确率更高,分类精度为76%。同时,第1主分量方差比重与核函数参数d之间满足递增关系。本文方法能够准确地提取火焰图像纹理特征。结论 本文提出一种对锅炉燃烧火焰图像进行状态识别的方法,对提取的火焰图像纹理特征向量进行降维并进行分类,可以获得较高的分类精度。实验结果表明,本文方法分类精度较高,运行时间较短,具有良好的实时性。     

Quantitative evaluation of a wavelet-based method in ventricular late potential detection

Ventricular late potentials (VLPs) are low-amplitude, high-frequency waveforms appearing in the terminal part of the QRS complex in electrocardiogram (ECG) of patients who are susceptible to ventricular tachycardia and sudden cardiac death, after surviving myocardial infarction. Accordingly, VLP detection presents a prominent non-invasive marker for some cardiac diseases clinically. This paper proposes a VLP detection method based on the wavelet transform and investigates its performance. In this method, a modified vector magnitude waveform is formed using discrete wavelet transform for each high-resolution ECG (HRECG) record; then, by applying the continuous wavelet transform to the QRS complex end part in this waveform, a feature vector is extracted from the resultant time-scale plot. This wavelet-based feature vector is processed by principle component analysis to reduce its dimensionality. Finally, a supervised feedforward artificial neural network, trained by a proper set of these feature vectors, is employed as a classifier. To evaluate the proposed method performance, a HRECG database consisting of the real VLP-negative and simulated VLP-positive patterns is used. In a comparative approach, different VLP detection techniques including the conventional time-domain method, developed by Simson, and some methods utilizing distinct diagnostic features are also applied to this database to investigate the capability of the proposed method in VLP analysis more completely. The results show the proposed method, employing the wavelet transform in both pre-processing and feature extraction stages, reveals high evaluation criteria (accuracy, sensitivity, and specificity) and is qualified to detect VLPs.     

基于多尺度分析的三维曲线匹配技术研究

从提取三维物体碎片轮廓曲线出发，提出一种基于多尺度分析的三维曲线匹配技术。轮廓曲线经多尺度平滑后，计算曲线的特征矢量，通过比较特征矢量以判断轮廓曲线的相似性，实现三维曲线匹配。实验表明提出的算法具有准确性、鲁棒性和容错性。     

基于Gabor小波与支持向量机对储粮害虫分类识别

将Gabor小波与支持向量机有机结合起来,对储粮害虫进行分类识别。首先对原始图像进行不同尺度和不同方向的Gabor滤波,获得原始图像特征,然后将遗传算法与支持向量机相结合来自动优选支持向量机模型参数,减少了以往应用中需反复实验来确定其参数的人工工作量。实验结果表明该方法识别率高,识别速度快,容错性好,而且能够正确识别有噪声的储粮害虫图像。为储粮害虫的快速鉴定和分类研究提供了可靠和科学的信息。     

基于奇异值分解和小波包分析的液压泵振动信号特征提取方法

针对液压泵故障特征提取问题,提出了一种基于奇异值分解和小波包变换的液压泵振动信号特征提取方法.通过奇异值分解将噪声非均匀分布的液压泵振动信号正交分解为噪声分布相对均匀的分量,对各分量进行小波包阈值去噪,重构去噪后分量,对去噪后信号进行小波包分解,提取各频带能量特征.以齿轮泵为例,将该方法对齿轮泵的气穴故障、齿轮磨损和侧板磨损3种常见故障和正常状态的振动信号进行特征提取分析,结果表明,该方法可有效提取齿轮泵故障特征.     

7500吨浮吊齿轮箱故障诊断系统的研究

针对7500吨浮吊齿轮箱故障诊断问题,将离散小波变换和Tikhonov支持向量机结合建立了一个浮吊齿轮箱故障诊断系统。在输入层对振动信号进行离散小波变换,提取不同频带的能量参数作为故障特征向量,利用这些特征向量进行Tikhonov支持向量机的学习,训练后的Tikhonov支持向量机诊断浮吊齿轮箱故障。实验结果表明,离散小波Tikhonov支持向量机具有很强的故障识别性能和鲁棒性,诊断精度优于常规的BP网络方法。     

Fault diagnosis and classification framework using multi-scale classification based on kernel Fisher discriminant analysis for chemical process system

Fault detection and diagnosis (FDD) in chemical process systems is an important tool for effective process monitoring to ensure the safety of a process. Multi-scale classification offers various advantages for monitoring chemical processes generally driven by events in different time and frequency domains. However, there are issues when dealing with highly interrelated, complex, and noisy databases with large dimensionality. Therefore, a new method for the FDD framework is proposed based on wavelet analysis, kernel Fisher discriminant analysis (KFDA), and support vector machine (SVM) classifiers. The main objective of this work was to combine the advantages of these tools to enhance the performance of the diagnosis on a chemical process system. Initially, a discrete wavelet transform (DWT) was applied to extract the dynamics of the process at different scales. The wavelet coefficients obtained during the analysis were reconstructed using the inverse discrete wavelet transform (IDWT) method, which were then fed into the KFDA to produce discriminant vectors. Finally, the discriminant vectors were used as inputs for the SVM classification task. The SVM classifiers were utilized to classify the feature sets extracted by the proposed method. The performance of the proposed multi-scale KFDA-SVM method for fault classification and diagnosis was analysed and compared using a simulated Tennessee Eastman process as a benchmark. The results showed the improvements of the proposed multiscale KFDA-SVM framework with an average 96.79% of classification accuracy over the multi-scale KFDA-GMM (84.94%), and the established independent component analysis-SVM method (95.78%) of the faults in the Tennessee Eastman process.