基于增强蜂群优化与k-means的文本聚类算法

针对文本数据维度较高、空间分布稀疏及其聚类效果不佳的问题,提出一种基于增强蜂群优化搜索与k-means的高效文本聚类算法。首先为蜂群算法引入公平操作与克隆操作来提高全局搜索的能力,公平操作提高了样本多样性并增强了蜂群搜索能力,克隆操作则增强了各代之间的信息交流,提高了求解质量。最终引入k-means进行局部质心的提炼,提高聚类质量。基于文本数据集的试验结果证明,相较于其他聚类算法,本算法具有更高的聚类质量。     

Data hiding in AMBTC images using quantization level modification and perturbation technique

Multimedia Tools and Applications - A novel data hiding method for Absolute Moment Block Truncation Coding (AMBTC) compressed image based on quantization level modification is proposed. Blocks of...     

Hypercolumn-array based image representation and its application to shape-based object detection

Biological and psychological evidence increasingly reveals that high-level geometrical and topological features are the keys to shape-based object recognition in the brain. Attracted by the excellent performance of neural visual systems, we simulate the mechanism of hypercolumns in the mammalian cortical area V1 that selectively responds to oriented bar stimuli. We design an orderly-arranged hypercolumn array to extract and represent linear or near-linear stimuli in an image. Each unit of this array covers stimuli of various orientations in a small area, and multiple units together produce a low-dimensional vector to describe shape. Based on the neighborhood of units in the array, we construct a graph whose node represents a short line segment with a certain position and slope. Therefore, a contour segment in the image can be represented with a route in this graph. The graph converts an image, comprised of typically unstructured raw data, into structured and semantic-enriched data. We search along the routes in the graph and compare them with a shape template for object detection. The graph greatly upgrades the level of image representation, remarkably reduces the load of combinations, significantly improves the efficiency of object searching, and facilitates the intervening of high-level knowledge. This work provides a systematic infrastructure for shape-based object recognition.     

A procedure for the generation of interval type-2 membership functions from data

This paper proposes a new interval type-2 fuzzy set taking extended π interval type-2 membership function (IT2 MF) as its values, and presents a new procedure for generating a set of extended π IT2 MFs from data for an interval type-2 linguistic variable. An extended π IT2 MF is defined as the min and max of two extended π (type-1 or ordinary) membership functions. The procedure has the following steps: (i) for each interval type-2 linguistic variable, specifying the number of membership functions to be generated, i.e. the granularity level, (ii) choosing two fuzzy exponents to be used, (iii) for each fuzzy exponent, applying the fuzzy c-means variant (FCMV) proposed by Liao et al. [1] to obtain the corresponding centers and membership values, and (iv) carrying out parametric optimization by applying a metaheuristic or a hybrid metaheuristic algorithm to determine the optimal parameters associated with the extended π IT2 MFs so that the mean squared error (MSE) or sum of squared errors (SSE) between the membership values obtained by FCMV and those predicted by the extended π IT2 MFs is minimized. The proposed procedure was illustrated with an example and further tested with iris data and weld data. The effects of using two different interval distance measures and the cluster means obtained by the FCMV as part of the initial solutions in the differential evolution metaheuristic were also investigated and discussed.     

Dynamics of quantum entanglement in quantum channels

Based on the von Neumann entropy, we give a computational formalism of the quantum entanglement dynamics in quantum channels, which can be applied to a general finite systems coupled with their environments in quantum channels. The quantum entanglement is invariant in the decoupled local unitary quantum channel, but it is variant in the non-local coupled unitary quantum channel. The numerical investigation for two examples, two-qubit and two-qutrit models, indicates that the quantum entanglement evolution in the quantum non-local coupling channel oscillates with the coupling strength and time, and depends on the quantum entanglement of the initial state. It implies that quantum information loses or gains when the state of systems evolves in the quantum non-local coupling channel.     

汽车客运站出入口监控系统中行人检测与跟踪方法

准确检测并跟踪行人目标是进行人流量统计和行为分析的基础。在汽车客运站出入口监控系统中,多个行人一起行走的情况较多。针对这种情况,本文提出一种能分割多人的行人检测与跟踪方法。首先通过背景差法提取运动行人,并对多人融合的情况下进行行人分割;然后结合卡尔曼滤波原理跟踪行人,并输出其运动轨迹。实验结果表明,该方法能清楚地绘出行人的运动轨迹,具有很好的鲁棒性。     

Curvature estimation for meshes based on vertex normal triangles

The estimation of surface curvature is essential for a variety of applications in computer graphics because of its invariance with respect to rigid transformations. In this article, we describe a curvature estimation method for meshes by converting each planar triangular facet into a curved patch using the vertex positions and the normals of three vertices of each triangle. Our method interpolates three end points and the corresponding normal vectors of each triangle to construct a curved patch. Then, we compute the per triangle curvature of the neighboring triangles of a mesh point of interest. Similar to estimating per vertex normal from the adjacent per triangle normal, we compute the per vertex curvature by taking a weighted average of per triangle curvature. Through some examples, we demonstrate that our method is efficient and its accuracy is comparable to that of the existing methods.     

Simulation-based two-stage estimation for multiple nonparametric regression

To reduce the curse of dimensionality arising from nonparametric estimation procedures for multiple nonparametric regression, in this paper we suggest a simulation-based two-stage estimation. We first introduce a simulation-based method to decompose the multiple nonparametric regression into two parts. The first part can be estimated with the parametric convergence rate and the second part is small enough so that it can be approximated by orthogonal basis functions with a small trade-off parameter. Then the linear combination of the first and second step estimators results in a two-stage estimator for the multiple regression function. Our method does not need any specified structural assumption on the regression function and it is proved that the newly proposed estimation is always consistent even if the trade-off parameter is designed to be small. Thus when the common nonparametric estimator such as local linear smoothing collapses because of the curse of dimensionality, our estimator still works well.     

智能规划中的可纳子目标排序

提出了一种称为可纳子目标排序(admissible subgoal ordering,简称ASO)的排序关系,给出了可纳排序的形式化定义并讨论其对增量式规划的重要性.随后介绍了原子依赖关系理论和原子依赖图技术,能够在多项式时间内近似求解可纳子目标排序关系.最后给出了一种计算可纳子目标序列的算法.其所有思想已经在规划系统ASOP中实现.通过在国际规划大赛标准测试领域问题上的实验,其结果表明,该方法能够有效地求解大规模的规划问题,并能极大地改善规划性能.     

基于Hot-Target 图和特征边缘保持的图像收缩方法

图像收缩是缩小高分辨率图像以适应不同纵横比小尺寸显示屏幕的过程,关键是收缩后能够凸显图像重要区域,保持连续,避免扭曲.提出一种新的图像收缩方法,该方法首先基于能量失真约束,迭代收缩覆盖图像的四边形网格至目标大小,然后映射,插值目标网格实现图像收缩.能量失真反映了对重要区域的凸显程度、结构的保持效果以及扭曲避免情况,失真越小,目标图像越理想.在该约束下,构成网格的子四边形非均匀收缩,重要度大的收缩小.为准确计算子四边形的重要度,根据图像显著度和边缘构建反映图像重要度的Hot-Target图.最后,通过保持图像直线边,称为特征边缘,避免非均匀收缩引起的边缘扭曲.为提高效率,降低复杂度,该方法由迭代求解线性方程实现.实验结果验证了方法的有效性.