基于模式复杂度的深度视频快速宏块模式选择算法

为了降低深度视频编码的复杂度,提高深度视频编码的效率,本文利用深度视频纹理复杂度的一致性,提出了一种深度视频快速宏块模式选择算法.该算法首先对不同区域模式复杂度进行分析,定义一个宏块模式复杂度因子,根据该因子将深度视频划分为简单模式区域和复杂模式区域.然后,分析深度视频的宏块模式分布情况及宏块模式选择的相关特性,为提出的算法提供理论依据.最后,根据模式区域选择相应的编码策略完成深度视频宏块模式的快速选择.实验结果表明,提出的算法在几乎不影响编码率失真性能及保证绘制质量的情况下可明显降低计算复杂度,能节约65.57％～92.72％的编码时间;而在低码率的情况下节约的编码时间会更多.因此,该算法在较低网络带宽限制下更加有效.

基于二维模式复杂度的海底地形分类方法

针对海底地形复杂程度分类问题,通过确定典型海底地形分类库中各类地形的二维模式复杂度,并以该二维模式复杂度值作为分类标准,将待确定海底地形的二维模式复杂度值与之比较,从而确定待分类地形的地形复杂程度.通过实测数据验证了基于二维模式复杂度的海底地形分类方法的有效性.

基于二维模式复杂度的DEM地形复杂研究

阐述了已有各地形复杂度描述因子的基础上,提出基于二维模式复杂度的栅格DEM地形复杂度因子的计算方法。计算陕西丘陵地区1∶5万DEM的结果表明,该方法可以有效地描述地表复杂地形的分布情况。

基于丙泊酚麻醉的相位模式复杂度同步及脑网络变化

麻醉诱导的无意识在宏观尺度表现为大脑区域之间功能连接的中断.然而,在介观尺度,麻醉的脑意识消失作用机制仍不明确.本研究分析了9名接受丙泊酚全身麻醉的癫痫患者(年龄18~54岁)的皮层脑电(electrocorticogram,ECoG)数据,采用了相位模式复杂度(phase-pattern complexity,PPC)的同步估计方法,并基于图论度量了皮层尺度麻醉前后的聚类系数、平均特征路径长度、全局效率、局部效率和中介中心性等几种网络特征参数.结果表明,delta-theta(1~8Hz)频段和gamma2(48~60Hz)频段在意识消失后的PPC值显著降低(分别为P<0.01,P<0.05).皮层尺度脑网络则由节点间的强连接转变为以少数节点为枢纽节点,并成散射状连接其他节点的网络拓扑结构.研究表明,麻醉导致的意识消失与皮层尺度神经活动的相位模式复杂度的降低及对应的拓扑网络的简单化有关.

多视点视频编码快速帧间模式选择算法

为解决多视点视频编码(MVC)可变尺寸块的模式选择计算复杂度过大的问题,提出了基于模式复杂度的多视点视频编码快速模式选择算法。该算法首先分析了多视点视频编码参考模型(JMVC)中各尺寸块的分布特点。然后,提出模式复杂度的概念以确定当前宏块的模式特征。最后,将宏块分成3个不同的模式类型:如果当前宏块属于简单模式,那么仅16×16分块被检查,其他分块均跳过;若当前宏块属于中等模式,则8×8块被跳过;若当前宏块属于复杂模式,所有模式分块都要检查。这样算法对那些不必要的模式选择过程就可以提前终止,从而使得计算量大大减少。实验结果表明:所提算法能保持同JMVC中全搜索算法几乎相同的编码效率,同时使计算复杂度减少62.75%。

基于多时-空相关的AVS2快速帧间预测算法

针对新一代视频标准AVS2引进四叉树分割、多参考帧等技术而带来的帧间预测复杂度增加的问题,提出一种基于多时-空相关的快速帧间预测算法。该算法利用上下层相邻编码单元(Coding Unit,CU)和空时域相邻CU在预测模式选择上的相关性,计算当前CU的模式复杂度,根据复杂度为当前CU分配合适的候选预测模式;同时利用相邻预测单元(Prediction Unit,PU)在参考帧选择上的相关性,计算当前PU的参考帧索引,根据索引为当前PU分配合适的候选参考帧。实验表明,该算法在BD-Rate(Bjontegaard delta bit rate)增加1.22%,BD-PSNR(Bjontegaard delta peak signal-to-noise rate)降低0.04 d B的前提下,平均减少47.54%的编码时间。

Low-Complexity Secure Network Coding Against Wiretapping Using Intra/Inter-generation Coding

Existing solutions against wiretapping attacks for network coding either bring significant bandwidth overhead or incur a high computational complexity.In order to reduce the security overhead of the existing solutions for securing network coding,a novel securing network coding paradigm is presented relying on two coding models:intra-generation coding and inter-generation coding.The basic idea to secure network coding using intra-generation coding is to limit the encryption operations for each generation,and then subject the scrambled and the remaining original source vectors to a linear transformation.This method is then generalized seamlessly using inter-generation coding by further exploiting the algebraic structure of network coding.We show that the proposed schemes have properties of low-complexity security,little bandwidth consumption,and high efficiency in integrating with the existing security techniques effectively.

Prediction of Planing Craft Motion Based on Grey System Theory

In order to minimize the harm caused by the instability of a planing craft, a motion prediction model is essential. This paper analyzed the feasibility of using an MGM（1,N） model in grey system theory to predict planing craft motion and carried out the numerical simulation experiment. According to the characteristics of planing craft motion, a recurrence formula was proposed of the parameter matrix of an MGMfl,N） model. Using this formula, data can be updated in real-time without increasing computational complexity significantly. The results of numerical simulation show that using an MGM（1,N） model to predict planing motion is feasible and useful for prediction. So the method proposed in this study can reflect the planing craft motion mechanism successfully, and has rational and effective functions of forecasting and analyzing trends.

Investigation of Combination Characteristic of Damage Location and Intensity in Arch Using Wavelet Coefficient

Arch is a typical complex structure comparing with beam and plate in bridge system. This paper investigates the damage characteristic combining the crack location with the crack intensity in arch. Initially, the first four displacement modes of intact and different damaged arch are simulated and the displacement mode changes are obtained. Next, the wavelet transformation is applied to the displacement mode changes in arch and wavelet coefficients at damage loci are picked. Finally, the damage index including damage location and damage intensity in arch is provided and plotted. The results show that wavelet coefficient module maximum of mode changes can be the damage indicator and is influenced by damage location and damage intensity. The damage indicator is proportional to the damage intensity and present monotonic trend according to damage location which depend on the mode order. At the same time, the large modulus maximum corresponds to small damage combination of location and intensity in the first four modes.

Pressure Wave and Fluid Velocity in a Bend-Mode Inkjet Nozzle with Double PZT Actuators

A new concept of bend-mode inkjet nozzle with double PZT (lead zirconate titanate) actuators has been designed and fabricated in the present study. Then the pressure wave and fluid velocity at the nozzle exit have been investigated. The complex pressure behavior inside the channel was solved numerically based on the narrow channel acoustic theory. The two PZTs attached to a rectangular channel were actuated sequentially by setting the waveforms of each PZT to be center-aligned with various pulse widths. As a result, the double PZT actuation is superior to the single PZT actuation in view of strong momentum force and fast dissipation of residual pressure. The maximum fluid velocity at the nozzle exit is observed when the respective pulse widths equal to their optimum pulse widths. The numerical results are supported by the experimental results with the fabricated inkjet device by measuring the speed of meniscus just out of the nozzle.

The Factors Analysis and Algorithm Implementation of Single-pattern Matching

By studying the algorithms of single pattern matching, five factors that have effect on time complexity of the algorithm are analyzed. The five factors are： sorting the characters of pattern string in an increasing order of using frequency, utilizing already-matched pattern suffix information, utilizing already-matched pattern prefix information, utilizing the position factor which is absorbed from quick search algorithm, and utilizing the continue-skip idea which is originally proposed by this paper. Combining all the five factors, a new single pattern matching algorithm is implemented. It＇s proven by the experiment that the efficiency of new algorithm is the best of all algorithms.