Discrimination of similar handwritten numerals based on invariant curvature features

This paper studies the discrimination of similar handwritten numerals based on invariant curvature features. High-order B-splines are used to calculate the curvature of the contours of handwritten numerals. The concept of a distribution center is introduced so that a one-dimensional periodic signal can be normalized as shift invariant. Consequently, the curvature of the contour of a character becomes rotation invariant. To reduce the dimension of the features, wavelet basis decomposition is used to produce more compact features. Finally, artificial neural network (ANN) and support vector machines (SVM) are employed to train the features and design classifiers of high recognition rates.     

Adaptive normalization of handwritten characters using global/localaffine transformation

This paper introduces an adaptive or category-dependent normalization method that normalizes an input pattern against each reference pattern using global/local affine transformation (GAT/LAT) in a hierarchical manner as a general deformation model. Also, the normalization criterion is clearly defined as minimization of the mean of nearest-neighbor interpoint distances between each reference pattern and a normalized input pattern. Optimal GAT/LAT is determined by iterative application of weighted least-squares fitting techniques. Experiments using input patterns of 3,171 character categories, including Kanji, Kana, and alphanumerics, written by 36 people in the cursive style against square-style reference patterns show that the proposed method not only can absorb a fairly large amount of handwriting fluctuation within the same category, but the discrimination ability is greatly improved by the suppression of excessive normalization against similarly shaped but different categories. Furthermore, comparative results obtained by the conventional shape normalization method for preprocessing are presented     

Pair-wise discrimination based on a stroke importance measure

The pair-wise discriminator is a binary classifier that verifies the outcome of the recognizer if it belongs to a class in a pre-defined confusion pair database. It is difficult to discriminate a pair of characters that are very similar in shape except for a small difference, because the small difference can be overridden by the writing variation. This paper proposes a pair-wise discrimination method that discriminates similar characters by focusing on the structural difference between the two characters. It discriminates a pair of characters by comparing their matching scores between the input character and the models of the two characters. When the stroke matching scores are combined to compute the overall matching score, each stroke is assigned a weight to reflect its importance in discriminating the character pair. By assigning large weights to the discriminative strokes, the difference between the characters is emphasized. The stroke weights are systematically obtained by a neural network training algorithm. In the experiments, the recognition performance was significantly improved by applying the proposed method.     

基于加权方法的节点重要性度量

网络中节点重要性度量对于信息的扩散、产品的曝光、传染性疾病的检测等都具有重大的理论意义。为了度量节点的重要性,基于网络拓扑结构考虑全局信息和局部信息提出了加权的节点重要性度量方法。具体的,对于一个无权网络,先考虑网络全局信息,算出每个节点的特征中心向量值,将边两端节点值的和作为边的权重,从而构成一个加权网络。然后根据加权网络的局部信息,求出加权网络的度。基于SIR模型的四个实证网络,实验结果表明加权方法比特征向量中心性、度中心性、紧密度中心性和介数中心性方法的效果更显著。     

Feature based robust watermarking using image normalization

This paper presents a robust watermarking scheme based on feature point detection and image normalization. Firstly some stable feature points are detected from the original image using the proposed multiresolution feature point detection filter. Then, image normalization is applied to the disks centered at these feature points. The watermark is embedded in the subband coefficients of DFT domain of each disk separately. And the watermark detection uses the correlation between the watermark embedding coefficients and the original watermark, and does not need the original image. The proposed scheme combines the advantages of feature point detection and image normalization, which can achieve strong robustness to signal processing and geometrical distortions. The experimental results also demonstrate good performance of the proposed scheme.     

基于模块化分析的动态系统重要度测量方法

为了解决关键系统灵敏性分析中的状态空间爆炸问题,提出一种基于模块化分析、适用于动态系统的重要度测量方法.将系统故障树分解为多个层次不同的独立模块,在此基础上引入微扰法的思想,对BI(Birnbaum importance measure)重要度指标进行扩展.提出相对重要度的概念,根据各独立模块的类型,分别求解相对重要度,利用相对重要度进行组合计算,求出系统级重要度.将该方法应用于一个实际高可靠计算机系统,分析结果显示该方法能够提供一个可信的重要度测量.     

一种基于汉字笔画的文本零水印算法

针对基于汉字的文本嵌入比较困难,水印不可见性不高,鲁棒性不高等特点,提出一种基于汉字笔画数的文本零水印新算法,算法通过统计各汉字在文本中的使用频率,即可得到使用频率最高的汉字,使用频率最高汉字的笔画数作为文本特征序列,用此序列与水印信息运算生成一组注册码。实验结果表明,该算法实现简单,具有很好的不可见性和鲁棒性。     

基于负载流的点加权复杂网络节点重要性评估方法研究

基于目前节点重要度评估方法大多针对静态无权网络拓扑结构的研究现状,考虑实际加权网络节点之间负载流动情况,从事理层面提出了一种基于网络贡献度的节点重要性评估方法。该方法基于典型加权复杂网络拓扑结构,从节点间负载流动和网络系统运行的角度出发,定义了流出负载量、流入负载量和流经负载量三个特征参数,并给出了具有一定现实意义的评价方法。该方法有助于更准确地发现复杂网络中的关键节点。最后的实验分析验证了该方法的有效性和可行性。     

Controller design based on similar skew‐symmetric structure for nonlinear plants

This paper proposes a novel controller design approach for nonlinear plants. A class of stable nonlinear systems with a similar skew‐symmetric structure is chosen as the objective closed loop system, and two design methods are proposed with backstepping and direct construction. Compared with the conventional backstepping method, the proposed backstepping method need not construct a Lyapunov function step by step, thus the design procedure is simplified. The direct construction method can be applied to some nonlinear plants for which the conventional backstepping is not feasible; and the design can be accomplished in only one step. Furthermore, for some nonlinear plants which have a lower triangular structure with two subsystems, simpler controllers can be derived by the proposed direct construction method than those derived by backstepping design. In addition, the proposed methods are both system structure oriented, therefore their designs are more intuitive than the conventional backstepping design. Two controllers are derived for satellite attitude control by employing the proposed methods; simulation results demonstrate their effectiveness. Copyright © 2009 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society     

Reliability evaluation and component importance measure for manufacturing systems based on failure losses

Little work has been done to assess the reliability of a vital system like the manufacturing system. In this article, a novel and effective system reliability evaluation method in terms of failure losses has been proposed for manufacturing systems of job shop type, and then the failure losses based component importance measure (CIM) is used for importance analysis of equipment. The former indicates the present system reliability situation and the latter points the way to reliability improvement efforts. In this scheme, the problem is described and modeled by a dynamic directed network. Consider that the actual processing time of machines is to contribute to failure occurrence, it is used to calculate the failure times and failure losses. The obtained total failure times and failure losses of the system are applied to evaluate its reliability. Techniques to estimate two kinds of failure losses based CIMs are presented. They offer guidelines to realize system reliability growth cost-effectively. A case study of a real job shop is provided as an example to demonstrate the validity of the proposed methods. Comparison to other commonly used methods shows the efficiency of the proposed methods.     

基于场势模型的网络节点重要性综合度量方法

针对网络节点的重要性受到多种因素影响的问题,提出一种基于场势模型的网络节点重要性综合度量方法。采用数据场模型对网络节点之间的相互作用关系进行形式化描述;根据节点的资产价值和互联节点之间的相互作用关系,综合量化节点的重要性;引入安全影响因素,通过可达性矩阵修正节点势场的可作用范围。理论分析与实验结果表明,该方法能够更客观准确地度量节点的重要性,有效区分边缘节点的差异性,使度量结果与节点的实际重要性相吻合。     

Efficient regional multi feature similarity measure based emotion detection system in web portal using artificial neural network

Emotion detection from facial expression has been well studied. There are numerous techniques has been discussed for the accuracy of emotion detection, however the methods suffer with higher false classification ratio. Towards the development of emotion detection, a novel region based multi feature similarity approach has been presented in this article. Considering, shape and geometry measure alone would not acquire higher performance in the classification. It is necessary to consider and combine multiple features towards the problem. With this motivation, the proposed Regional Multi Feature Similarity (RMFS) based emotion detection algorithm enhances the input facial image and extracts shape feature, geometry feature and wrinkle features with colors are considered. Extracted features are trained with neural network. At the classification stage, MFS measure has been estimated towards the features of various emotion class in different layers of neural network. Finally, a single one has been classified as result using artificial neural network. The proposed method improves the performance of emotion detection with reduced false ratio.     

反网络钓鱼中UNICODE字符相似度评估算法的研究

针对网络钓鱼以URI（Uniform Resource Identifier）及IRI（Internationalized Resource Identifier）域名地址欺骗的主要特点,研究组成相似域名的字符相似度评估。在目前一些研究成果的基础上,提出了基于字符点阵重叠新型的快速字符相似度评估算法,并基于理论方法设计算法设计相似字符评估原型软件,得出改进前后算法的实验结果,证明了方法得到优化,从而为今后建立相似域名库工作做好准备。     

非线性Contourlet变换和区域特征匹配的图像融合

针对图像融合中存在边缘轮廓表示不清晰和计算冗余度过高的问题.提出了将非线性估计Contourlet变换和图像区域特征相结合的融合算法.通过非线性估计保留Contourlet高频稀疏矩阵最重要系数频带,对低频区域、高频区域采用不同的区域特征策略进行图像融合.实验结果表明:该图像融合算法不仅可以提高图像融合的视觉效果和图像质量、而且可以去噪音,增强图像.     

基于矩阵测度的非线性广义时滞系统鲁棒模糊控制

研究T-S模糊广义时滞系统的鲁棒控制问题.不同于传统的寻求公共正定矩阵的方法,基于矩阵测度给出保证系统鲁棒稳定的充分条件,并将此条件进一步转化为线性矩阵不等式.通过求解线性矩阵不等式,得到状态反馈控制器和静态输出反馈控制器.最后通过算例仿真验证了方法的有效性.     

Texture segmentation using independent-scale component-wise Riemannian-covariance Gaussian mixture model in KL measure based multi-scale nonlinear structure tensor space

This paper proposes a novel texture segmentation approach using independent-scale component-wise Riemannian-covariance Gaussian mixture model (ICRGMM) in Kullback-Leibler (KL) measure based multi-scale nonlinear structure tensor (MSNST) space. We use the independent-scale distribution and full-covariance structure to replace the covariant-scale distribution and 1D-variance structure used in our previous research. To construct the optimal full-covariance structure, we define the full-covariance on KL, Euclidean, log-Euclidean, and Riemannian gradient mappings, and compare their performances. The comparison experiments demonstrate that the Riemannian gradient mapping leads to its optimum properties over other choices when constructing the full-covariance. To estimate and update the statistical parameters more accurately, the component-wise expectation-maximization for mixtures (CEM²) algorithm is proposed instead of the originally used K-means algorithm. The superiority of the proposed ICRGMM has been demonstrated based on texture clustering and Graph Cuts based texture segmentation using a large number of synthesis texture images and real natural scene textured images, and further analyzed in terms of error ratio and modified F-measure, respectively.     

基于字符规范化的文档图像水印算法

运用字符规范化和小波变换的知识,提出一种将文档图像分割成字符图像,再对字符图像规范化,然后将随机序列嵌入到小波图像低频系数的水印算法.根据视觉系统纹理掩蔽特性,将不同强度的水印分量嵌入到了不同的小波系数中.由于文档图像分割和规范化本身具有抗几何攻击的特性,故该方法对缩放、小角度旋转有一定的鲁棒性,实验结果表明:该方法在文档图像上比其它方法更具备优越性.     

基于透射率归一化的单幅图像去雾算法

在有雾天气条件下拍摄的图像,由于光线在传播人过程中受到空气中悬浮颗粒的散射,导致图像内容模糊不清,颜色偏灰白色。为了恢复出清晰的图像,根据大气散射物理模型,提出一种单幅图像去雾算法。首先,该算法从大气散射模型出发,对大气光模型进行变形化简,得到新的去雾模型。然后对大气光值估计,并利用在局部统计上不相关性恢复透射率。最后,对透射率进行归一化处理,带入模型,得到恢复图像。实验结果表明,该算法能够很好地恢复图像的对比度和保持图像的真实颜色。     

Efficient numerical simulation methods for estimating fuzzy failure probability based importance measure indices

For the problem with a fuzzy failure state which commonly exists in degradation structures and systems, the fuzzy failure probability based importance measure indices can be used to measure the effect of the input variables on the fuzzy failure probability effectively. However, the computational cost is unaffordable for estimating the indices directly. For efficiently evaluating the fuzzy failure probability based importance measure indices, this paper proposed two numerical simulation methods, i.e., the direct Monte Carlo simulation method based on the Bayesian formula (B-DMCS) and the adaptive radial-based importance sampling method based on the Bayesian formula (B-ARBIS). The two proposed methods employ the Bayesian formula to eliminate the dependence of the computational cost on the dimensionality of the input variables. Compared with the B-DMCS method, the B-ARBIS method can enhance the computational efficiency significantly due to repeatedly utilizing the same group of samples of the input variables and the strategy to adaptively search the optimal hypersphere in the safety domain. After giving the principles and implementations of the two methods, three examples are employed to validate the effectiveness of the two proposed numerical simulation methods. The results of the examples demonstrate that the effectiveness of the two proposed methods is higher than the direct Monte Carlo method, and the B-ARBIS method can improve the efficiency obviously in contrast with the B-DMCS method.

     

A quadtree normalization scheme based on cyclic translations

An image can be represented by a compact and hierarchical structure, i.e. a quadtree. The storage requirements of an image constructed by a quadtree is highly sensitive to its position. Ang and Samet [Pattern Recognition Lett. 15(1), 57–63 (1994)] proposed an algorithm capable of normalizing a quadtree in O(s²log₂s) time and O(s²) space, where s is the length of the image grid, such that the number of nodes of the quadtree after normalization can be minimal. However, s is twice as long as the length of one side of the image to be normalized. In this study, we propose a normalization scheme based on cyclic translations. The time complexity and the space requirements of this scheme have four times less than those in Ang and Samet's case. In addition, no translation is necessary to fit the image into the northwest quadrant of the grid before the process of normalization. Also, this scheme can normalize a quadtree to obtain less node numbers than that of Ang and Samet. Furthermore, if the image's four corners have the same color, the amount of reduction for node number becomes larger after cyclic translations; it can occasionally reach to 75%. The analytical and empirical results demonstrate the advantages of this scheme.