A novel video abstraction method based on fast clustering of the regions of interest in key frames

Online video nowadays has become one of the top activities for users and has become easy to access. In the meantime, how to manage such huge amount of video data and retrieve them efficiently has become a big issue. In this article, we propose a novel method for video abstraction based on fast clustering of the regions of interest (ROIs). Firstly, the key-frames in each shot are extracted using the average histogram algorithm. Secondly, the saliency and edge maps are generated from each key-frame. According to these two maps, the key points for the visual attention model can be determined. Meanwhile, in order to expand the regions surrounding the key points, several thresholds are calculated from the corresponding key-frame. Thirdly, based on the key points and thresholds, several regions of interest are expanded and thus the main content in each frame is obtained. Finally, the fast clustering method is performed on the key frames by utilizing their ROIs. The performance and effectiveness of the proposed video abstraction algorithm is demonstrated by several experimental results.     

Simulating ion clustering in potassium thiocyanate aqueous solutions with various ion-water models

Using a molecular dynamics simulation technique,we compared several commonly used ion-water models to describe the microscopic structures and dynamics in KSCN aqueous solutions.Results are compared with observations of femtosecond infrared vibrational-energy transfer and anisotropy measurements.The Jorgensen/TIP4P model is found to provide the best reproduction of clustering properties such as percentage of clustered ions,cluster-size distribution,concentration dependence of the water,and ion-rotation time constants.     

多集值信息表中的多粒度模糊决策论粗糙集

多粒度粗糙集和决策论粗糙集是Pawlak粗糙集的重要推广,目前已成为人工智能研究的热点.然而,它们大多处理的都是单值信息系统中的问题.而实际生活中绝大多数都是处理多值问题,为了解决这一问题,在多集值信息表中将多粒粗糙集与模糊决策论粗糙集相结合进行研究,提出了其在乐观,悲观情形下的上下近似,研究了一些相关性质并给出了多集值信息表中的多粒度模糊决策论粗糙集精度、粗度的概念,最后通过一个具体例子验证其有效性.     

MOEPGA: A novel method to detect protein complexes in yeast protein–protein interaction networks based on MultiObjective Evolutionary Programming Genetic Algorithm

The identification of protein complexes in protein–protein interaction (PPI) networks has greatly advanced our understanding of biological organisms. Existing computational methods to detect protein complexes are usually based on specific network topological properties of PPI networks. However, due to the inherent complexity of the network structures, the identification of protein complexes may not be fully addressed by using single network topological property. In this study, we propose a novel MultiObjective Evolutionary Programming Genetic Algorithm (MOEPGA) which integrates multiple network topological features to detect biologically meaningful protein complexes. Our approach first systematically analyzes the multiobjective problem in terms of identifying protein complexes from PPI networks, and then constructs the objective function of the iterative algorithm based on three common topological properties of protein complexes from the benchmark dataset, finally we describe our algorithm, which mainly consists of three steps, population initialization, subgraph mutation and subgraph selection operation. To show the utility of our method, we compared MOEPGA with several state-of-the-art algorithms on two yeast PPI datasets. The experiment results demonstrate that the proposed method can not only find more protein complexes but also achieve higher accuracy in terms of fscore. Moreover, our approach can cover a certain number of proteins in the input PPI network in terms of the normalized clustering score. Taken together, our method can serve as a powerful framework to detect protein complexes in yeast PPI networks, thereby facilitating the identification of the underlying biological functions.     

基于模式识别轮轴识别器的改良与应用

轮轴识别器模式识别的改良与应用，适用于高速公路计重收费中，采用模式识别技术，在统计分析大量货车计重数据的基础上，结合现有车型建立模型数据库，根据计重仪表提供的货车轴间距、轴重、车长等数据，分析判断得出货车轴型。克服了传统的轮轴识别器在使用过程中由于外界的泥沙、油污、冻冰现象导致的识别率下降，维修率逐年提高，通过软件技术和单片机电子技术成功解决了这一难题。     

模糊层次综合评判法在电梯风险评估中的应用

针对电梯系统设备部件老化、故障逐步上升,电梯常规定期检验预测电梯事故和潜在故障作用有限的现状,为了提高电梯的安全使用,通过综合电梯系统主观和客观因素,运用模糊理论中层次综合评判法,建立了电梯系统的安全状态综合评判方法。首先说明了该方法的原理和建立流程,然后结合电梯轿厢的安全评估,给出了评估案例。结果表明,该方法简单易用,为电梯系统的安全评判提供了技术支持。     

基于FPGA的电机伺服器设计

文中针对传统的8位单片机由于计算功能等条件限制和系统内部体系结构,在实现各种高效的控制算法和先进的电机控制理论时面临的着成本高,系统复杂程度高等问题.该设计采用模糊比例算法,即在大范围内采用模糊控制,以提高系统的动态响应速度,建立系统模型.通过现场可编程门阵列FPGA的数字电子系统对电机进行控制,以硬件电路算法实现算法程序,将原来的电路板级产品集成为芯片级产品,为电动机数字控制提供了一种新的有效方法.该设计的优点:低功耗,可靠性高,低成本.     

面向领域标签辅助的服务聚类方法

Web服务数量的激增对服务发现提出了更高的要求,服务聚类是促进服务发现的一种重要技术.但是,现有服务聚类方法只对单一类型的服务文档进行聚类,缺乏考虑服务的领域特性和服务标签的应用.针对这些问题,本文首先使用本体辅助的支持向量机和面向领域的服务特征降维技术建立服务的特征内容向量,然后使用一种标签辅助的主题服务聚类方法T-LDA建立融合标签信息之后的隐含主题表示,并利用归一化方法消除通用主题的影响,综合上述方法建立一个面向领域标签辅助的Web服务聚类方法DTWSC.实验结果表明,该框架能够提高针对不同类型的服务文档的聚类效果.与LDA、K-Means等方法相比,该方法在聚类纯度、熵和F-Measure指标上均具有更好的效果.     

一种改进的基于KNN颜色线性模型抠图算法*

介绍了一种减少用户标记和改进的基于KNN(K Nearest Neighbors)颜色线性模型的图像软抠取算法.通过ESCG(Efficient Spectral Clustering on Graphs)算法对输入图像进行谱聚类,用户只需选择某些类中确定的前景、背景像素,便能生成只包含少数未知像素的三分图.基于KNN颜色线性模型的抠图算法将局部平滑假设与非局部原理相结合,但在毛发及前景背景像素近似区域抠取效果并不理想,提出的改进算法将焦点特征添加到特征向量中,最小化基于图拉普拉斯矩阵的二次目标函数并确定泰知像素的透明度值.实验表明,改进算法在毛发、孔洞或者图像前景背景近似的区域都能有好的抠取效果.