ID 原文 译文
5794 在跟踪的过程中,根据当前帧的颜色直方图基于贝叶斯分类器建立空间注意机制,在提取VGGNet19中多层卷积特征后分别与空间注意图进行融合,从而构建稳健的目标表观模型。 In the process of tracking, according to the current frame color histogram based on bayesian classifier set up the mechanism of spatial attention, after multiple convolution characteristics in extracting VGGNet19 respectively the integration and spatial attention to figure, to build the steady apparent model.
5795 之后利用学习到的相关滤波器得到各响应值,通过加权求和准则求出最终响应,同时利用帧差法调整学习速率,最终实现自适应的目标跟踪。 After learning to utilization of filter response value is obtained, by weighted summation criterion and the final response, at the same time frame differential method is used to adjust the learning rate, finally realizes the adaptive target tracking.
5796 实验结果表明,所提算法在大多数复杂环境下的跟踪准确度和鲁棒性均优于现有的跟踪算法。 The experimental results show that the proposed algorithm under the complex environment of most of the tracking accuracy and robustness are better than the existing tracking algorithm.
5797 线性预测是时间序列分析中常用方法,针对传统一维线性预测谱估计算法只能估计信号源角度或信号频率问题,提出空时二维线性预测算法。 Linear prediction is commonly used in time series analysis method, in view of the traditional one dimensional linear prediction spectral Angle estimation algorithm can estimate signal source and signal frequency, space-time two-dimensional linear prediction algorithm.
5798 采取对空时二维阵列接收到的数据进行数据抽取和排列,和对数据协方差矩阵进行重新构造的方法,求取空时二维线性预测权值并进行谱峰搜索。 Take the space-time two-dimensional array receives the data selection and arrangement of data, and to reconstruct the data covariance matrix, the method of calculating weights of space-time two-dimensional linear prediction and spectral peak searching.
5799 重点分析了空时二维的前向预测、后向预测和双向预测算法的原理,着重研究了构造的空时二维线性预测协方差矩阵的数据结构。 Analyzed the space-time two-dimensional to predict before and after the forecast and the principle of bi-directional prediction algorithm, this paper studies the structure of the space-time two-dimensional linear prediction covariance matrix data structure.
5800 讨论了前向、后向和双向预测的相互关系以及二维与一维的关系,并与空时二维最小方差算法、空时二维多重信号分类(multiple signal classification,MUSIC)算法进行了对比与分析。 Discussed the forward and backward and bi-directional prediction correlation, and the relationship between the two-dimensional and a one-dimensional and space-time two-dimensional minimum variance algorithm, space-time two-dimensional multiple signal classification (multiple signal classification, MUSIC) algorithm are compared and analyzed.
5801 理论分析与仿真表明,一维空域、一维时域算法的前向、后向和双向预测为空时二维预测算法的特例。 Theoretical analysis and simulation indicate that the one-dimensional airspace, one dimensional time domain algorithm of forward and backward and bi-directional prediction for a special case of the space-time two-dimensional prediction algorithm.
5802 同时空时二维预测算法不仅克服了空时二维最小方差算法、空时二维MUSIC算法不能解相干信号源的缺点,还具有很好的测向测频能力。 At the same time the space-time two-dimensional prediction algorithm not only overcomes the space-time two-dimensional minimum variance algorithm, space-time 2-d MUSIC algorithm can't coherent signal source of faults, but also has good direction frequency measurement capability.
5803 为对抗无线传感器网络中存在的强针对性的智能型干扰攻击,基于多种扩频技术提出了一种混合跳频扩频、非协调直接序列扩频以及非协调跳频扩频的抗干扰攻击方法。 To fight that exist in the wireless sensor network (WSN) strong attack targeted intelligent interference, a hybrid was developed based on the many kinds of spread spectrum technology, the coordination of direct sequence spread spectrum and frequency hopping spread spectrum of coordination of frequency hopping spread spectrum anti-interference attack methods.