针对时间序列多步预测的聚类隐马尔科夫模型

时间序列的预测在现今社会各个领域中有着广泛的应用.本文针对时间序列趋势预测中的多步预测问题,提出了基于聚类的隐马尔科夫模型,利用隐马尔科夫模型中的隐状态来表示产生时间序列数据时的系统内部状态,实现对多步时间序列的预测.针对时间序列聚类中的距离计算问题,提出结合时间序列时间性和相似性的聚类算法,并给出了迭代精化基于聚类的隐马尔科夫模型的方法.实验表明,本文提出的方法在时间序列多步预测中精度较高.     

基于N阶近邻路网的车辆行程时间估计模型

车联网的提出为智能交通的研究提供了新的交通信息收集技术.针对短时交通中车辆的路网行程时间估计问题,提出了基于N阶近邻的隐马尔科夫模型,利用马尔科夫性质来解决道路行程时间的前后关联性问题,同时考虑不同道路的异构性构建了N阶近邻路网模型来模拟路网间的交互影响.针对短时交通中实时数据更新的问题,提出基于道路关联性算法,并结合车联网的采集技术给出了迭代更新模型的方法.实验表明,本文提出的方法在短时交通车辆行程时间预测中精度较高,能够在车辆行进中做出实时预测.     

基于不同隐马尔科夫模型的图像识别方法

为了更详细地研究隐马尔科夫模型在图像识别中的应用,以指纹识别为例,纵向总结了几种基于隐马尔科夫模型的指纹图像识别算法,包括一维隐马尔科夫模型、伪二维隐马尔科夫模型、二维模型及一维模型组。分别从时间复杂度、识别精确度等方面总结出这四种隐马尔科夫模型在图像识别时的优缺点,得出不同待识别图像适合使用的识别模型的结论。     

基于WSN的隐马尔科夫模型煤矿井下定位

针对煤矿井下定位精度不高的问题,提出一种隐马尔科夫模型(Hidden Markov Model,HMM)的定位方法。通过比较隐马尔科夫模型和粒子滤波模型推断出,当定位的数据需要通过多跳的方式传播时,使用半马尔科夫平滑移动模型(Semi-Markov Smooth Mobility Model,SMS)来估计误差,能量消耗,能提升定位精度。     

基于多维正交载体的可裂解流指纹方案

针对流交换中流源身份不可知、流交换范围不可控和流路径不可追踪问题,提出了基于多维正交载体的可裂解流指纹方案。利用2种相互正交的载体提高指纹信息的容量,并通过时间间隔重心载体特性和基于隐马尔科夫模型的解码技术实现可裂解性,提高方案的健壮性。分析了基于重心属性值随机选取载体的反制多流攻击能力,以及不同条件下指纹重心标记算法和基于隐马尔科夫模型解码技术的正确率。最后通过实验对算法的健壮性和隐蔽性进行了探讨。     

基于复小波邻域隐马尔科夫模型的图像去噪

多分辨信号和图像模型可用于捕获图像中平滑和奇异区域的统计结构,但是,基于正交小波变换的模型受到平移变化的影响从而降低了其准确性和实时性.本文将邻域隐马尔科夫模型LCHMM( Local Contextual Hidden Markov Model)扩展到复小波的范围,提出了一种基于复小波的邻域隐马尔科夫模型C-LCHMM( Local Contextual Hidden Markov Model Based On Complex Wavelet),该模型具有近似平移不变性及分辨率高的特点、能够捕获小波系数的邻域的统计特征、且计算复杂度小.仿真试验表明基于复小波邻域隐马尔科夫模型(C-LCHMM)用于图像去噪的效果优于典型的去噪算法.     

扩频通信与扩频调制技术

本文提出一种用于 CDMA 扩频系统中窄带干扰抑制的基于隐马尔科夫模型的新颖非线性估计算法。所提出的算法组合了一个递归隐马尔科夫模型估计器,一个卡尔曼滤波器和一个递归预测误差参数估计算法。文中说明了所提出的算法的性能不仅优于用于窄带干扰抑制的非线性滤波技术,而且也优于基于交叉耦合期望最大化的算法,并且有更快、更顽健的收敛特性。参17     

基于隐马尔科夫模型的P2P流识别技术

为了实时、准确地识别多种P2P应用流,提出了基于隐马尔科夫模型(HMM,hidden Markov model)的P2P流识别技术。该技术利用分组大小、到达时间间隔和到达顺序等特征构建流识别模型,采用离散型随机变量刻画HMM状态特征;提出了能同时识别多种P2P应用流的架构HMM-FIA,设计了HMM的状态个数选择算法。在校园网中架设可控实验环境,使用HMM-FIA识别多种P2P流,并与已有识别方法进行比较,结果表明采用离散型随机变量能降低模型建立时间,提高识别未知流的实时性和准确性;HMM-FIA能同时识别多种P2P协议产生的分组流,并能较好地适应网络环境变化。     

基于隐马尔科夫模型的波动率预测探究

基于沪深300指数收益率,利用隐马尔科夫模型对不可直接观测的波动率进行研究。通过历史数据进行参数估计,然后运用参数和当前观测值对未来波动率进行预测。之后,将预测结果与GARCH模型的预测结果进行比较,实证结果表明了隐马尔科夫模型的有效性和准确性。     

基于HMM的加权Textrank单文档的关键词抽取算法

文中介绍了现有几种比较流行的关键词提取技术,提出了基于隐马尔科夫模型的加权Textrank的单文档关键词抽取算法。对比分析了三种算法的效果:基于词频的关键词提取算法,基于词性、位置、频度的关键词提取算法,加权Textrank算法。实验结果表明加权Textrank算法在单文档提取中有较好的效果,并且在单篇文章提取较少的关键词时准确率较高。     

Segmentation and tracking in echocardiographic sequences: activecontours guided by optical flow estimates

This paper presents a method for segmentation and tracking of cardiac structures in ultrasound image sequences. The developed algorithm is based on the active contour framework. This approach requires initial placement of the contour close to the desired position in the image, usually an object outline. Best contour shape and position are then calculated, assuming that at this configuration a global energy function, associated with a contour, attains its minimum. Active contours can be used for tracking by selecting a solution from a previous frame as an initial position in a present frame. Such an approach, however, fails for large displacements of the object of interest. This paper presents a technique that incorporates the information on pixel velocities (optical flow) into the estimate of initial contour to enable tracking of fast-moving objects. The algorithm was tested on several ultrasound image sequences, each covering one complete cardiac cycle. The contour successfully tracked boundaries of mitral valve leaflets, aortic root and endocardial borders of the left ventricle. The algorithm-generated outlines were compared against manual tracings by expert physicians. The automated method resulted in contours that were within the boundaries of intraobserver variability     

基于变形模板技术的目标轮廓跟踪算法

为了适应目标旋转、尺度、场景光照等变化,利用B样条曲线表达目标轮廓,结合变形模板技术对运动目标的轮廓进行跟踪.在跟踪过程中,沿着模板曲线的法线方向检测目标轮廓,提高了检测效率.根据同一曲线上相邻点间的相关性,对检测所得的轮廓点集的坐标序列进行中值滤波,有效降低噪声干扰.将检测到的轮廓点集在形状空间匹配,使目标轮廓的形变限制在一定范围之内,有效抑制噪声和背景边缘特征的干扰.仿真试验表明,该算法能够有效得到目标轮廓,且具有较好的实时性.     

Automated optic disk boundary detection by modified active contour model

This paper presents a novel deformable-model-based algorithm for fully automated detection of optic disk boundary in fundus images. The proposed method improves and extends the original snake (deforming-only technique) in two aspects: clustering and smoothing update. The contour points are first self-separated into edge-point group or uncertain-point group by clustering after each deformation, and these contour points are then updated by different criteria based on different groups. The updating process combines both the local and global information of the contour to achieve the balance of contour stability and accuracy. The modifications make the proposed algorithm more accurate and robust to blood vessel occlusions, noises, ill-defined edges and fuzzy contour shapes. The comparative results show that the proposed method can estimate the disk boundaries of 100 test images closer to the groundtruth, as measured by mean distance to closest point (MDCP) <3 pixels, with the better success rate when compared to those obtained by gradient vector flow snake (GVF-snake) and modified active shape models (ASM).     

一种快速超宽带穿墙雷达成像算法

超宽带穿墙雷达因为反恐、灾后救援等方面的迫切需要而成为近年来的一个研究重点。传统的超宽带成像算法在目标识别能力及计算效率上已不能满足其要求。基于逆边界散射变换的SEABED成像算法可以完成对目标边界快速成像,但是墙体的存在使其在穿墙雷达应用中受到限制。该文考虑到墙体的影响,在SEABED算法基础上提出一种新的适用于超宽带穿墙雷达的包络线算法,该算法通过计算从回波中提取的准波前所对应的曲线簇的包络线来完成目标边界成像。利用FDTD数值仿真实现了对隐藏在墙后目标的二维成像,验证了算法的有效性,并分析了墙壁对目标定位误差的影响。     

Three-dimensional statistical model for gingival contour reconstruction

Optimal gingival contours around restored teeth and implants are of critical importance for restorative success and esthetics. This paper describes a novel computer-aided methodology for building a 3-D statistical model of gingival contours from a 3-D scan dental dataset and reconstructing missing gingival contours in partially edentulous patients. The gingival boundaries were first obtained from the 3-D dental model through a discrete curvature analysis and shortest path searching algorithm. Based on the gingival shape differential characteristics, the boundaries were demarcated to construct the gingival contour of each individual tooth. Through B-spline curve approximation to each gingival contour, the control points of the B-spline curves are used as the shape vector for training the model. Statistical analysis results demonstrate that the method can give a simple but compact model that effectively capture the most important variations in arch width and shape as well as gingival morphology and position. Within this statistical model, the morphologically plausible missing contours can be inferred based on a nonlinear optimization fitting from the global similarity transformation, the model shape deformation and a Mahalanobis prior. The reconstruction performance is evaluated through large simulated experimental data and a real patient case, which demonstrates the effectiveness of this approach.     

DWT BASED HMM FOR FACE RECOGNITION

A novel Discrete Wavelet Transform (DWT) based Hidden Markov Module (HMM) for face recognition is presented in this letter. To improve the accuracy of HMM based face recognition algorithm, DWT is used to replace Discrete Cosine Transform (DCT) for observation sequence ex- traction. Extensive experiments are conducted on two public databases and the results show that the proposed method can improve the accuracy significantly, especially when the face database is large and only few training images are available.     

基于流形特征与形状先验的红外直升机图像分割

为有效分割复杂天空背景下的直升机目标,提出了基于流形特征与形状先验的变分分割模型.根据图像数据的灰度分布构造区域能量项,推动变形曲线向目标边界演化;引入对称正定（Symmetric Positive Definite,SPD）矩阵流形上的区域协方差描述子构造流形特征能量项以提高分割算法的鲁棒性.在区域项、边界项和流形特征项的共同作用下获取红外直升机目标的第一阶段分割结果.在第二阶段分割过程中,基于主成分分析（Principal Component Analysis,PCA）方法获取直升机目标的先验形状变化模式,以变形曲线在PCA空间重构的形状作为先验知识约束曲线的演化,最终实现红外直升机图像的分割.实验结果表明,本文方法能够有效获取直升机目标的完整轮廓.     

HMM转移概率的新的重估算法

将隐含马尔可夫模型(HMM:Hidden Markov Model)引入到语音识别中来是一个巨大的贡献.但是在经典的HMM中关于状态转移概率a_ij(#em/em#≠j)与自转移概率a_ii的独立性假设,导致了这个模型的不协调性.事实上,段长分布概率与状态转移概率并非相互独立的,由其中的一个就可以唯一的确定另外一个.本文从段长分布概率出发说明了以上关于转移概率独立性假设的不合理性,并得到了转移概率新的重估算法.这个新算法比经典HMM的Baum-Welch迭代算法重估转移概率效果更好,前者比后者相对误识率下降了大约5%.     

网络最大流问题的一种新的符号ADD求解算法

通过对网络及网络最大流问题的符号代数判定图描述,在Trff算法的基础上,给出了网络最大流问题的一种新的符号ADD求解算法。与Dinic、Karzanov算法相比,本文算法的空间复杂度较低。实验结果表明,该算法可处理更大规模的问题。     

Global regularizing flows with topology preservation for active contours and polygons.

Active contour and active polygon models have been used widely for image segmentation. In some applications, the topology of the object(s) to be detected from an image is known a priori, despite a complex unknown geometry, and it is important that the active contour or polygon maintain the desired topology. In this work, we construct a novel geometric flow that can be added to image-based evolutions of active contours and polygons in order to preserve the topology of the initial contour or polygon. We emphasize that, unlike other methods for topology preservation, the proposed geometric flow continually adjusts the geometry of the original evolution in a gradual and graceful manner so as to prevent a topology change long before the curve or polygon becomes close to topology change. The flow also serves as a global regularity term for the evolving contour, and has smoothness properties similar to curvature flow. These properties of gradually adjusting the original flow and global regularization prevent geometrical inaccuracies common with simple discrete topology preservation schemes. The proposed topology preserving geometric flow is the gradient flow arising from an energy that is based on electrostatic principles. The evolution of a single point on the contour depends on all other points of the contour, which is different from traditional curve evolutions in the computer vision literature.