一种基于分层的人类动作识别和定位算法研究

人类动作识别在视频自动分析、视频检索等领域获得广泛应用,是目前的研究热点。然而现有的动作识别方法重点关注视频的非静态部分而忽略大部分静态部分,从而影响了动作识别和定位的效果。本文提出一种新的分层空间-时间分段表示法,以分层方式实现部位和整个身体的多分辨率表示,可用于运动识别和定位。该算法分为3个步骤。第一步,首先对每个视频帧进行分层分段,以得到一组分段树,每颗树是身体分段树的候选。第二步,利用视频的轮廓、接合对象结构、全局前景色等信息对候选分段树进行修剪。第三步,在时域上对剩余分段层的每个分段进行前向和后向跟踪。我们以难度较大的UCF-Sports和HighFive数据集为实验对象,对本文方法进行性能评估,实验结果表明,本文方法的性能要优于当前最新运动检测算法性能,运动定位性能与当前最新算法相当。     

Spanning 3-ended trees in k-connected K 1,4-free graphs

A tree with at most m leaves is called an m-ended tree.Kyaw proved that every connected K1,4-free graph withσ4(G)n-1 contains a spanning 3-ended tree.In this paper we obtain a result for k-connected K1,4-free graphs with k 2.Let G be a k-connected K1,4-free graph of order n with k 2.Ifσk+3(G)n+2k-2,then G contains a spanning 3-ended tree.     

PTreeRec: Phylogenetic Tree Reconstruction based on genome BLAST distance

Phylogenetic Tree Reconstruction (PTreeRec) is a web-based tool for automatic phylogeny inferences from whole-genome sequences, which accepts files of DNA sequences in the FASTA format and allows users to save the output tree file, and displays the inferred tree through an applet in a web browser. PTreeRec involves three basic steps. First, regions of maximal segment pairs (MSPs) based on an all-against-all pairwise comparison of genomes are located. Second, a distance matrix is calculated from MSP scores or coverage. Finally, a phylogenetic tree is reconstructed by the neighbor-joining method.     

Integro-local and integral theorems for sums of random variables with semiexponential distributions

We obtain some integro-local and integral limit theorems for the sums S(n) = ξ(1) + ? + ξ(n) of independent random variables with general semiexponential distribution (i.e., a distribution whose right tail has the form $P(\xi \ge t) = e^{ - t^\beta L(t)} $ , where β ∈ (0, 1) and L(t) is a slowly varying function with some smoothness properties). These theorems describe the asymptotic behavior as x → ∞ of the probabilities P(S(n) ∈ [x, x + Δ)) and P(S(n) ≥ x) in the zone of normal deviations and all zones of large deviations of x: in the Cramér and intermediate zones, and also in the “extreme” zone where the distribution of S(n) is approximated by that of the maximal summand.     

Large deviations of the first passage time for a random walk with semiexponentially distributed jumps

Suppose that ξ, ξ(1), ξ(2), ... are independent identically distributed random variables such that ?ξ is semiexponential; i.e., $P( - \xi \geqslant t) = e^{ - t^\beta L(t)} $ is a slowly varying function as t → ∞ possessing some smoothness properties. Let E ξ = 0, D ξ = 1, and S(k) = ξ(1) + ? + ξ(k). Given d > 0, define the first upcrossing time η ₊(u) = inf{k ≥ 1: S(k) + kd > u} at nonnegative level u ≥ 0 of the walk S(k) + kd with positive drift d > 0. We prove that, under general conditions, the following relation is valid for $u = (n) \in \left[ {0, dn - N_n \sqrt n } \right]$ : 0.1 $P(\eta + (u) > n) \sim \frac{{E\eta + (u)}}{n}P(S(n) \leqslant x) as n \to \infty $ , where x = u ? nd < 0 and an arbitrary fixed sequence N _n not exceeding $d\sqrt n $ tends to ∞. The conditions under which we prove (0.1) coincide exactly with the conditions under which the asymptotic behavior of the probability P(S(n) ≤ x) for $x \leqslant - \sqrt n $ was found in [1] (for $x \in \left[ { - \sqrt n ,0} \right]$ it follows from the central limit theorem).     

External estimation of a segment function by a polynomial strip

The problem is considered of constructing a least-width strip with a polynomial axis that contains the graph of a given continuous segment function. Convex analysis methods are used to obtain a criterion for solving the problem in a form comparable to the Chebyshev alternance. Sufficient conditions for the uniqueness of a solution are given, including those taking into account the differential properties of the segment function to be estimated.     

基于需求分类和乘客升级购买行为的航空公司收益管理模型

研究了基于乘客分类的航空客运库存控制与动态定价策略.模型中,航空公司以提供折扣票的方式将乘客分为两类,并针对购买折扣票的乘客存在升级购买行为,通过动态的控制折扣票的销售和对机票实施动态定价来最大化自身的期望收益.应用动态规划建立了相应的收益管理模型,讨论了最优定价应满足的关系式,并得到了接受或拒绝乘客购买折扣票的阈值.最后,通过算例分析了升级购买概率对阈值、机票的价格及期望收益的影响.     

嵌段聚酯型聚脲氨酯弹性体的红外光谱研究

用红外光谱法对基于4,4′－二苯基甲烷二异氰酸酯（MDI）、乙二胺（ED）和聚己二酸丁二醇酯（PBA）（分子量分别为1975和1228）的嵌段聚脲氨酸（PUU）弹性体进行了研究。首先归属了PUU的红外吸收谱带,并对前人已经归属过的1600、1317、1520和1230cm^-1谱带提出了新的看法。氚化和变温技术的使用为上述红外谱带的归属和氢键化研究提供了可靠的理论依据。研究结果表明氢键化主要发生在硬段聚集区中,少量的氢键在软、硬段之间的界面上形成,由此推断PUU材料是相分离的。     

氘化聚酯型聚脲氨酯弹性体的红外光谱研究

用红外光谱法对基于4,4′－二苯基甲烷二异氰酸酯（MDI）、乙二胺（ED）和聚己二酸丁二醇酯（PBA）的嵌段聚酯型聚脲氨酯（PUU）弹性体及其氚化产物进行了研究。结果表明氚化反应对PUU中涉及PUU中涉及氢键化N－H振动红外吸收光谱带的影响较大,受氚化影响较大的那些谱带分别在3332,3190,1600,1539,1317,1257和1230cm^-1附近。另外也发现不同硬段含量和不同聚酯软段分子量的PUU受氚化影响的程度基本相似。     

基于矩和小波变换的目标图象识别

在图象目标的识别中,目标图象的不变矩对于具有旋转、尺度变换特性的目标具有良好特征匹配性能,但其识别准确度直接受到不变矩计算区域内的目标个数的影响.本文首先采用一维小波变换降噪并获取图象边缘,结合轮廓跟踪法(“爬虫法”)与投影法对图象进行区域分割获取各目标存在区域,再利用目标图象的不变矩进行目标识别,最后通过利用目标的主轴特性及质心变化计算目标的旋转角度及尺度变换倍率.通过仿真实验,证明方案可行且具有良好的识别性能.