基于Matlab平台人脸识别仿真过程的探究

Matlab是一种科学与工程计算的高级语言,广泛地运用在包括信号与图像处理、控制系统设计等方面,我们使用Matlab作为平台,设计GUI,探究人脸识别的过程。在特征提取方面,运用主成分分析法(PCA)算法,对高维特征进行降维,保证了高位数据不失真,在分类器算法上采用支持向量机(SVM)和自适应提升(Adaboost)算法进行对比实验,SVM通过求解由全部训练样本对检测样本最佳线性表示的稀疏向量来进行分类,Adaboost算法针对不同的训练集训练同一个基本分类器(弱分类器)进行多次迭代,每次迭代增加错样本的权重,构成一个更强的最终的分类器(强分类器),实验结果表明,仿真能够达到较高的识别率和缩短识别的时间。     

一种SVM集成的图像分类方法研究*

针对单分类器没有充分考虑数据集的特征而不能很好地完成分类识别,提出了一种基于集成学习技术的SVM集成的图像分类方法。该方法是在基于较为流行的词袋(Bag-of-Words,BOW)模型的图像分类方法的基础上,利用训练生成的不同SVM分类器分类测试图像,并将分类结果采用集成学习算法进行集成。分别采用传统的BOW模型的图像分类方法和本文提出的方法进行分类实验,实验结果表明采用SVM集成的图像分类方法明显提高了分类精度,具有一定的稳健性。     

基于联合特征和SVM的极化SAR图像分类方法

极化SAR地物分类作为极化SAR数据解译的关键环节,已成为遥感领域研究的一个新热点。在充分研究现有方法的基础上,给出了一种联合特征和SVM相结合的极化SAR图像分类方法。该方法基于目标分解理论提取极化SAR图像的多类散射特征,并结合具有上下文知识的纹理特征,构建联合特征矢量;利用提取样本区域像素的联合特征矢量训练SVM分类器;将未知数据输入训练好的分类器完成最终的分类。实测SAR图像数据的实验结果表明,算法能够充分利用极化SAR图像电磁散射特性及纹理特征的互补性,具有较好的分类性能。     

一种基于权重融合的JPEG隐写分析方法

针对JPEG图像通用隐写检测中检测效率低、训练时间长的问题,提出一种基于集成分类器的新检测方法。算法以CC-PEV为特征对图像进行描述并作为隐写分析特征;然后,随机构造若干个特征子空间,用bootstrap方法构造图像训练子集,分别进行训练得到数个基分类器;根据基分类器的分类结果赋予基分类器不同的权重,将基分类器的结果按照其权重进行融合得到最终的结果。本文对该算法进行了测试,对它的集成性、检测准确率和训练时间进行分析。实验结果表明,相对于传统的集成方法,本文方法用自举方法构造训练集、随机方法构造子特征空间、赋予基分类器不同权重进行融合能够显著地提高算法准确率。本文方法相对于SVM和传统的集成分类方法,具有更高的检测率,对于特征维数更大的图像检测,具有更好的拓展性和一般适用性。     

一种基于Hadoop平台CloudSVM的网络流量分类方法

大规模的netflow训练数据集是构建高质量、高稳定网络流量分类器的必然要求。但随着网络流特征维数的提高和数据集规模的扩大,无论是网络流的分析处理还是基于支持向量机(SVM)的分类器模型的训练,都无法在有效的时间内得到有效的处理结果。本文基于Hadoop云计算平台,采用MapReduce技术对SVM网络流量分类器进行分布式学习和训练,构建CloudSVM网络流量分类器。通过对来自校园网出口镜像的近2 T的大规模网络流量的跟踪文件的分布式存储和处理,对抽取的样本数据集进行分类,实验验证了基于Hadoop平台分布式存储和并行处理大规模网络数据集的高效率性,也验证了CloudSVM分类器在不降低分类准确度的情况下可以快速收敛到最佳,并随着大规模网络流样本的增加,SVM分类器训练的时间趋近平稳。     

多分类器融合的快速高维特征聚类图像分割

提出一种多分类器融合的快速高维特征聚类图像分割方法,将图像高维特征数据的分类分解为基于灰度(颜色)特征的最佳模糊分类以及基于空域约束的统计分类等多个低维特征数据的分类.通过多分类器融合的方法将不同分类器得到的分类结果进行优化整合,得到最后的分类结果.实验证明:与其它图像分类算法相比,该方法拥有更好的分割性能并大大提高了计算速度,最大限度地保证了分割算法计算的简单有效性.     

基于感兴趣区域和HOG-MBLBP特征的交通标识检测

交通标识检测中样本类别间的不平衡常常导致分类器的检测性能弱化,为了克服这一问题,该文提出一种基于感兴趣区域和HOG-MBLBP融合特征的交通标识检测方法。首先采用颜色增强技术分割提取出自然背景中交通标识所在的感兴趣区域;然后对标识样本库提取HOG-MBLBP融合特征,并用遗传算法对SVM交叉验证进行参数的优化选取,以此来训练和提升SVM分类器性能;最后将提取的感兴趣区域图像的HOG-MBLBP特征送入训练好的SVM多分类器,进行进一步的精确检测和定位,剔除误检区域。在自建的中国交通标识样本库上进行了实验,结果表明所提方法能达到99.2%的分类准确度,混淆矩阵结果也表明了该方法的优越性。     

基于迁移学习和SVM的糖网图像分类

为解决基于卷积神经网络(CNN)的糖尿病视网膜病变(DR)图像分类算法上普遍存在模型参数难以训练、易过拟合的问题,本文提出一种基于迁移学习和支持向量机(SVM)分类器的DR图像分类算法。首先,对DR图像进行预处理和数据扩增;其次,采用迁移学习方法预初始化深度学习分类算法中的经典框架VGGNet-16网络的模型参数,固定浅层网络参数不变,微调深层网络参数;最后,提取VGGNet-16最后一个隐藏层的特征向量训练支持向量机(SVM)分类器判定DR图像是否病变。实验结果表明,在Kaggle-DR公共数据集共35126张DR图像进行实验,在随机抽取的3500张作为测试集,分类准确率为0.931、敏感性为0.933、特异性为0.928,并能加快网络收敛和提高模型的泛化性。     

基于随机蕨的极化SAR图像地物分类研究

传统极化SAR图像地物分类方法通常存在计算效率低和维度灾难等问题，受益于随机蕨分类器的简单性、鲁棒性和处理高维特征空间的能力，文中提出了一种基于随机蕨算法的极化SAR分类框架算法。随机蕨分类器中大量的二元特征捕获了极化SAR图像中地物的空间信息、纹理属性和与其相邻像素的关系。该方法能够在人工标注像素数量较少的情形下对极化SAR图像进行准确、高效的地物分类并且所需要的训练一个随机蕨分类器的时间仅需几十秒。最终的分类实验结果表明，该方法在Oberpfaffenhofen数据集上达到了较好的分类性能和运行效率。     

基于改进型深度数据流形的数据分类算法及在人脸中的应用

针对数据分类问题的局限,提出一种基于改进型深度数据流形的数据分类算法并将其应用到人脸识别中。首先,通过采集人脸图像的深度信息,利用稀疏表示对其进行去噪处理;再结合图像的颜色信息,重新生成三维人脸信息数据库,通过对人脸数据的流形分析得到最优的降维结果,按十字十乘交叉验证法的原则选取训练集和测试集,将训练集输入支持向量机算法建立数据分类器;最后,将测试集输入训练完成的分类器中,实现人脸数据分类。选取ORL,Yale两类人脸图像标准数据库与传统人脸识别算法进行交叉对比实验,验证算法的优越性和可行性。实验结果表明：所提出的算法有较高的分类准确率,可有效地完成人脸识别。     

Tumor segmentation in lung CT images based on support vector machine and improved level set

In lung CT images, the edge of a tumor is frequently fuzzy because of the complex relationship between tumors and tissues, especially in cases that the tumor adheres to the chest and lung in the pathology area. This makes the tumor segmentation more difficult. In order to segment tumors in lung CT images accurately, a method based on support vector machine (SVM) and improved level set model is proposed. Firstly, the image is divided into several block units; then the texture, gray and shape features of each block are extracted to construct eigenvector and then the SVM classifier is trained to detect suspicious lung lesion areas; finally, the suspicious edge is extracted as the initial contour after optimizing lesion areas, and the complete tumor segmentation can be obtained by level set model modified with morphological gradient. Experimental results show that this method can efficiently and fast segment the tumors from complex lung CT images with higher accuracy.     

基于机器视觉的印刷套准识别方法研究*

针对印刷套准检测存在的精度低、速度慢的问题,提取了印刷标志图像的Tamura纹理特征：粗糙度、对比度和方向度,以描述其印刷标志套准或套不准特征;设计了支持向量机的分类器对印刷标志图像进行套准识别,并采用高斯径向基核函数用于非线性数据的分类。实验结果证明,采用建议的印刷标志图像特征提取和分类方法,识别准确率达到90%,识别时间为0.032751秒。本文建议的方法在识别准确率和识别速度上都优于人工检测和文献8的方法。     

Alzheimer's diagnosis using eigenbrains and support vector machines

An accurate and early diagnosis of the Alzheimer's disease (AD) is of fundamental importance for the patient's medical treatment. Single photon emission computed tomography (SPECT) images are commonly used by physicians to assist the diagnosis. Presented is a computer-assisted diagnosis tool based in a principal component analysis (PCA) dimensional reduction of the feature space approach and a support vector machine (SVM) classification method for improving the AD diagnosis accuracy by means of SPECT images. The most relevant image features were selected under a PCA compression, which diagonalises the covariance matrix, and the extracted information was used to train an SVM classifier, which could classify new subjects in an unsupervised manner.     

Study on the classification of multi-spectral images based on a FSVM multi-class classifier with the binary tree

Due to the features of the multi-spectral images, the result with the usual methods based on the support vector machine (SVM) and binary tree is not satisfactory. In this paper, a fuzzy SVM multi-class classifier with the binary tree is proposed for the classification of multi-spectral images. The experiment is conducted on a multi-spectral image with 6 bands which contains three classes of terrains. The experimental results show that this method can improve the segmentation accuracy. 更多还原     

基于高阶统计量的合成图像鉴别方法

随着各种图像编辑软件的广泛应用,数字图像被篡改的现象越来越普遍。论文研究了一种鉴别合成照片真伪的方法,在提取可信图像和拼接图像的双谱特征和边缘密度特征的基础上利用支持向量机(SVM)进行分类训练,并提出了在特征提取前去除图像的非高斯噪声。实验结果表明,该方法可将鉴别正确率由71%提高到75%,并具有一定的抗JPEG压缩能力。     

基于Bayesian粗糙集和布谷鸟算法的肺部肿瘤高维特征选择算法

在高维特征选择过程中最优特征子集生成和分类器 参数优化方面,提出一种基于贝叶斯粗糙集(BRS)、遗传算法(GA)和布谷鸟算法(CS) 的两阶段优化高维特征选择算法。该算法首先分析3000例肺部肿瘤CT图像的形状、灰度和纹理特征,提取104维特 征分量共同量化ROI;然后进行两阶段优化:(1) 从全局相对增益函数的角度分析了属性 重要度,结合属性约简长度和基因编码权值函数的加权和构造适应度函数,通过选择、交叉 和变异等遗传操作生成最优特征子集,在不降低分类精确度的前提下降低特征维度;(2) 利用CS对支持向量机(SVM)参数进行全局寻优;最后通过实验验证本文算法的可行性和有 效性。实验结果表明,该算法有效提升了肺部肿瘤良恶性识别能力,降低了算法的时间复杂 度。     

基于极化干涉SAR图像的地物监督分类方法

X波段的高分辨率极化干涉合成孔径雷达（synthetic aperture radar,SAR）图像包含较强的斑点噪声,不利于地物分类等应用.针对这一问题,先使用Nonlocal滤波进行预处理,然后提取图像的极化特征和干涉特征,再使用支持向量机（support vector machine,SVM）和AdaBoost分类器对极化和干涉特征矢量进行分类.利用N-SAR系统于渭南市采集的极化干涉SAR数据进行验证,该数据共包含10类地物,并有足够的ground truth用来进行分类器的训练和测试.实验结果表明,AdaBoost分类器能对多类地物取得较好的分类效果,且干涉信息的加入能带来一定改善.     

Comparison of ICA approaches for facial expression recognition

Independent component analysis (ICA) and Gabor wavelets extract the most discriminating features for facial action unit classification by employing either a cosine similarity measure (CSM) classifier or support vector machines (SVMs). So far, only the ICA approach, which is based on the InfoMax principle, has been tested for facial expression recognition. In this paper, in addition to the InfoMax approach, another five ICA approaches extract features from two facial expression databases. In particular, the Extended InfoMax ICA, the undercomplete ICA, and the nonlinear kernel-ICA approaches are exploited for facial expression representation for the first time. When applied to images, ICA treats the images as being mixtures of independent sources and decomposes them into an independent basis and the corresponding mixture coefficients. Two architectures for representing the images can be employed yielding either independent and sparse basis images or independent and sparse distributions of image representation coefficients. After feature extraction, facial expression classification is performed with the help of either a CSM classifier or an SVM classifier. A detailed comparative study is made with respect to the accuracy offered by each classifier. The correlation between the accuracy and the mutual information of independent components or the kurtosis is evaluated. Statistically significant correlations between the aforementioned quantities are identified. Several issues are addressed in the paper: (i) whether features having super- and sub-Gaussian distribution facilitate facial expression classification; (ii) whether a nonlinear mixture of independent sources improves the classification accuracy; and (iii) whether an increased “amount” of sparseness yields more accurate facial expression recognition. In addition, performance enhancements by employing leave-one-set of expressions-out and subspace selection are studied. Statistically significant differences in accuracy between classifiers using several feature extraction methods are also indicated.     

Combining feature scaling estimation with SVM classifier design using GA approach

This letter adopts a GA （Genetic Algorithm） approach to assist in learning scaling of features that are most favorable to SVM （Support Vector Machines） classifier, which is named as GA-SVM. The relevant coefficients of various features to the classification task, measured by real-valued scaling, are estimated efficiently by using GA. And GA exploits heavy-bias operator to promote sparsity in the scaling of features. There are many potential benefits of this method： Feature selection is performed by eliminating irrelevant features whose scaling is zero, an SVM classifier that has enhanced generalization ability can be learned simultaneously. Experimental comparisons using original SVM and GA-SVM demonstrate both economical feature selection and excellent classification accuracy on junk e-mail recognition problem and Internet ad recognition problem. The experimental results show that comparing with original SVM classifier, the number of support vector decreases significantly and better classification results are achieved based on GA-SVM. It also demonstrates that GA can provide a simple, general, and powerful framework for tuning parameters in optimal problem, which directly improves the recognition performance and recognition rate of SVM.