ID 原文 译文
14325 为了解决依靠光学传感器进行手势识别对外部环境依赖较大的问题，提出了一种基于连续波( Continuous Wave，CW) 雷达的手势识别方法，并建立了 4 种手势动作的回波数据库。 To solve the problem that external environment affects recognition accurary，a new hand gesturerecognition method based on continuous wave( CW) radar is investigated，and a dataset of four typical handgestures echo wave is generated．
14326 首先，对CW 雷达回波进行短时傅里叶变换( Short-Time Fourier Transform，STFT) 获取手势动作的时频谱。 Firstly，the short-time Fourier transform( STFT) of CW radar echoes fromhand gestures is applied to produce the time-frequency spectrograms of hand gestures．
14327 然后，通过设立阈值将时频谱中的背景杂波去除。 Next，the backgroundclutter in the time-frequency spectrograms is removed by setting up thresholds．
14328 接下来，对处理后的时频谱提取方向梯度直方图( Histogram of Oriented Gradient，HOG) 特征。 Then，the histogram of oriented gradient( HOG) feature is extracted from the processed time-frequency spectrograms．
14329 最后，采用支持向量机( Support Vector Machine，SVM) 作为分类器，以 HOG 特征作为输入进行手势识别。 Finally，the support vector machine( SVM) is used as a classifier and the HOG features are used as input for gesture recognition．
14330 实验结果表明，所提方法在普通室内环境下的识别精度超过 95% ，能够对典型的手势动作进行有效识别。 The experimental results show that the recognition accuracy of the proposed method reaches morethan 95% in a normal indoor environment，and it can effectively recognize typical hand gestures．
14331 针对传统通信扩频体制在统计学上具有较强的规律性而易被敌方侦测设备发现的缺点，提出了一种抗扩频参数测量的低截获波形方案。 For the shortcoming that general spread spectrum( SS) communication structure is easy to be intercepted by the enemy detection equipment due to its statistically regular signal characteristics，a low probability of intercept( LPI) waveform scheme with rejection capability of SS parameter measurements is proposed．
14332 从抑制和消除通信信号中存在的可检测、可攻击统计特征的思路出发，通过引入随机过程对跳码扩频序列和码速率进行随机化控制，使输出的码速率和扩频周期时间具有非平稳统计特征和非各态历经性质，以增强低截获波形抗扩频参数测量的能力。 From the perspective of eliminating the detectable signal statistical characteristics，the proposedwaveform introduces stochastic process to randomly control code hopping SS sequences and SS code rates toendow SS code rate and SS period with non-stationary statistical characteristics，greatly enhancing the rejection capability of LPI waveform in terms of SS parameter measurements．
14333 仿真结果表明，相比于固定增益定码扩频体制，所提出的低截获波形具备码速率和扩频周期时间测量精度低的特点，且在高信噪比条件下存在较高的相对稳态误差。 Simulation results indicate thatcompared with SS communication structure with specific SS sequences，the proposed LPI waveform has theadvantages of the low measurement accuracy of SS code rate and SS period，and maintains a quite high relative steady-state error．
14334 针对现存无线传感器网络定位算法中需要采集、存储和处理大量数据导致运算量较大与能耗过高的问题，提出了一种改进的基于贝叶斯压缩感知的多目标定位算法。 For the problem that a large amount of data is needed to be collected，stored and processed in theexisting localization algorithms for wireless sensor networks and the computation is large and energy consumption is high，an improved multi-stage localization algorithm based on Bayesian compressive sensing isproposed．