基于子空间分析的人脸识别方法研究

人脸识别技术是模式识别和机器视觉领域的一个重要研究方向,在众多人脸识别的算法中,基于子空间分析的特征提取方法以其稳定可靠的识别效果成为了人脸识别中特征提取的主流方法之一。本文对目前应用较多的子空间分析方法进行了研究,具体介绍了线性子空间分析方法：主成分分析（PCA）、线性鉴别分析（LDA）、独立主成分分析（ICA）、快速主成分分析（FastICA）等及非线性子空间分析方法：基于核的PCA （KPCA）等的基本思想及其在人脸识别中的研究进展,包括一些新的研究成果。此外,还应用orl及Yale B人脸库对几个基础的子空间方法进行了验证实验。实验结果表明,在几个子空间分析方法中,FastICA算法取得了最高的识别率。最后结合实验结果对各算法的优缺点进行了分析总结。     

用改进的基因算法求大气压强公式

提出了一种利用改进的基因算法求大气压强精确公式的参数反演识别方法,运用基因算法反演求出由多态方程推导所得大气压强公式中的大气比热比γ,从而得到大气压强精确计算公式,并与有关文献给出的大气压强公式进行比较。结果表明,用基因算法所确定的大气压强公式所得计算值与实测值更加吻合,说明基因算法在反演问题中具有很高的应用价值,同时经数据分析得出改进基因算法比传统基因算法精度高,速度快。     

Parameter identification of chaos system based on unknown parameter observer

Parameter identification of chaos system based on unknown parameter observer is discussed generally. Based on the work of Guan et al. [X.P. Guan, H.P. Peng, L.X. Li, et al., Acta Phys. Sinica 50 (2001) 26], the design of unknown parameter observer is improved. The application of the improved approach is extended greatly. The works in some literatures [X.P. Guan, H.P. Peng, L.X. Li, et al., Acta Phys. Sinica 50 (2001) 26; J.H. Lü, S.C. Zhang, Phys. Lett. A 286 (2001) 148; X.Q. Wu, J.A. Lu, Chaos Solitons Fractals 18 (2003) 721; J. Liu, S.H. Chen, J. Xie, Chaos Solitons Fractals 19 (2004) 533] are only the special cases of our Corollaries 1 and 2. Some observers for Lü system and a new chaos system are designed to test our improved method, and simulations results demonstrate the effectiveness and feasibility of the improved approach.     

基于阻抗模拟的等效电磁参数研究

为能够较好的计算混合介质等效复介电常数和复磁导率, 提出了一种基于等效电路理论的阻抗模拟方法.导出了材料的电磁参数与材料容抗、感抗之间的对应关系, 并建立了能够较完整、精确地刻画实际复合材料的模型.通过求解混合介质的等效阻抗, 进一步反演其等效电磁参数.将计算结果与经典理论公式以及基于有限差分的数值方法进行了比较, 结果都符合得很好, 说明该方法可以用于计算材料的等效电磁参数.另外, 引入薄膜层结构来刻画表面效应等因素, 使得计算结果更加符合实验结果, 弥补了经典公式在这方面的不足, 同时也体现了该方法在刻画实际模型时所具有的优势. 关键词： 等效电磁参数 阻抗模拟 等效电路 混合介质     

Controlling the Chen attractor using linear feedback based on parameter identification

In this paper we investigate the method for controlling the chaotic Chen system with unknown parameters using linear feedback functions based on parameter identification.The observer is applied to the identification of the unknown parameters of Chen system.Then an efficient feedback function is designed for controlling the Chen system.Numerical simulations are provided to show the effectiveness and feasibility of the developed controller.     

基于车载电子标签数据的单交叉口状态判别研究

基于目前交叉口状态判别存在设备安装复杂度高,判别准确率低,受环境影响大、无法识别具体车辆等问题,提出基于车载电子标签数据的交叉口状态判别方法。通过简化模型,以两相位单交叉口为例,选择的参数为流量,交通密度,停车数量。最后经VISSIM交通仿真软件的二次开发,模拟车载电子标签运行环境,通过对比仿真试验所得与专家观察分析所得,相似度高达90%以上。证明该方法在克服以往方法缺陷的基础上可有效判别交叉口状态。     

Krylov subspace iterative methods for boundary element method based near-field acoustic holography

The reconstruction of the acoustic field for general surfaces is obtained from the solution of a matrix system that results from a boundary integral equation discretized using boundary element methods. The solution to the resultant matrix system is obtained using iterative regularization methods that counteract the effect of noise on the measurements. These methods will not require the calculation of the singular value decomposition, which can be expensive when the matrix system is considerably large. Krylov subspace methods are iterative methods that have the phenomena known as "semi-convergence," i.e., the optimal regularization solution is obtained after a few iterations. If the iteration is not stopped, the method converges to a solution that generally is totally corrupted by errors on the measurements. For these methods the number of iterations play the role of the regularization parameter. We will focus our attention to the study of the regularizing properties from the Krylov subspace methods like conjugate gradients, least squares QR and the recently proposed Hybrid method. A discussion and comparison of the available stopping rules will be included. A vibrating plate is considered as an example to validate our results.     

Adaptive generalized projective synchronization in different chaotic systems based on parameter identification

In this Letter, the generalized projective synchronization of different chaotic systems with unknown parameters is investigated. By Lyapunov stability theory, the adaptive control method is proposed to achieve above synchronization phenomenon. Meanwhile, according to the invariance principle of differential equations, unknown parameter can be estimated accurately. The schemes are successfully applied to two groups of examples: the anti-phase synchronization between Lorenz system and Chen system; the complete synchronization between hyper-chaotic system and generalized Loren system. The corresponding numerical results are presented to verify the effectiveness of this method.     

基于频谱分析与支持向量机的车型音频识别研究

机动车车型识别是城市道路交通流监测统计的一个重要方面。本文基于频谱分析与支持向量机方法提出一种车型音频识别方法,以1/3倍频程频谱数据作为特征数据,并使用支持向量机方法完成不同车型分类下的车型识别,同时还分析比较了不同训练样本量及不同单个样本数据量大小对识别结果的影响。在将车型细分的情况下,对小汽车、大型公交车、水泥车、摩托车四种车型的样本外识别结果达到96.9%的准确率,验证了方法的有效性。     

Investigation into MIS structures based on gradedband-gap hetero-epitaxial HgCdTe grown by molecular-beam epitaxy using photo-emf and conductivity methods

The effect of graded-band-gap layers on the differential resistance of space-charge region in MIS structures based on MBE HgCdTe (x = 0.225) is examined. It is shown that the effect of resistance of the epitaxial-film bulk on the measured capacitance and resistance should be taken into account for correct determination of space-charge region parameters. The presence of near-surface layers with increased Cd contents results in an increase in the resistance of the space-charge region in strong inversion. The product of semiconductor resistance by area as high as 15 Ω⋅cm² is obtained despite suppression of tunnel generation-recombination through deep levels in MIS structures with graded-band-gap layers. This might be due to background photogeneration and diffusion of minority charge carriers. The mechanisms for limitation of the differential resistance of space-charge region at different temperatures are discussed for n-HgCdTe (x = 0.225 and 0.292) and p-HgCdTe (x = 0.225).     

Decimative subspace-based parameter estimation methods of magnetic resonance spectroscopy based on prior knowledge

The method Hankel Total Least Squares (HTLS)-PK, which successfully incorporates prior knowledge of known signal poles into the method HTLS, has been proven to greatly improve the performance for parameter estimation of overlapping peaks of magnetic resonance spectroscopy (MRS) signal. In addition, decimation is also proposed as a way to increase the performance of subspace-based parameter estimation methods in the case of oversampling. Taking advantage of decimation in combination with prior knowledge to estimate the MRS signal parameters, two novel subspace-based parameter estimation methods, HTLSDSumPK and HTLSDStackPK, are presented in this paper. The experimental results and relevant analysis show that the methods HTLSDSumPK, HTLSDStackPK and HTLS-PK are slightly better than the method HTLS at low noise levels; however, the three prior-knowledge-incorporating methods, especially the method HTLSDSumPK, have much better performance than the method HTLS at high noise levels in the terms of robustness, estimated accuracy and computational complexity. Even if some inaccuracy of prior knowledge is considered, the method HTLSDSumPK also shows some advantages.     

基于一阶时滞混沌系统参数辨识的保密通信方案

基于参数调制原理,针对一阶时滞Logistic混沌系统设计了相应的参数观测器,仅传递单路信号就可以成功辨识出系统中的未知参数,从而恢复出调制在参数中的模拟信号,并在此基础上采用不同频率信号代替“0”或“1”,设计出具体的数字保密通信方案.数值仿真验证了本方法的有效性. 关键词： 观测器 混沌保密通信 参数调制 滤波     

An unmanned aerial vehicle identification and tracking system based on weakly supervised semantic segmentation technology

This paper presents an unmanned aerial vehicle (UAV) identification and tracking system aimed at monitoring UAVs based on weakly supervised semantic segmentation. A camera is equipped with a pan–tilt to collect images for semantic segmentation network in real time. The GrabCut+ algorithm and annotation boxes are employed to generate the UAV “pseudo pixel labels” for supervised model learning and reduce labelling costs. A new loss function combining the focus loss function and dice loss function is designed to balance positive and negative samples and improve the segmentation effect. The Mixup method is introduced for model training to prevent overfitting and enhance the generalization ability of the model. The semantic segmentation network outputs the prediction results by a fully connected conditional random field to smooth the target image. Furthermore, a region-based tracking method is proposed to solve the hysteresis problem of the pan–tilt control system and improve the system tracking performance. Finally, an experiment based on a dataset is carried out to prove the effectiveness of the segmentation algorithm with 66.3% mIoU. Considering that 10% of the central area of view is specified as the view centre, a UAV falling in the centre of the field accounts for more than 80% of this view area, demonstrating the real-time effectiveness of the designed UAV identification and tracking system.     

基于无消相干子空间的量子避错码设计

针对量子系统的联合消相干模型,可以找到一些不受消相干错误影响的系统状态,这种状态被称为相干保持态,所有相干保持态构成的空间就称为无消相干子空间(decoherencefreesubspace,缩写为DFS).利用DFS的特性可以实现自动容错的量子避错码.首先提出一种DFS的定义,并且以定理的形式证明其他DFS的定义都是等价的.然后给出了DFS的存在性定理.最后利用群论的方法设计一种构造DFS的简单方法 关键词： 相干保持态 无消相干子空间 量子避错码 容错量子计算     

Adaptive synchronization and lag synchronization of uncertain dynamical system with time delay based on parameter identification

In this paper, the adaptive synchronization and lag synchronization are considered for uncertain dynamical system with time delay based on parameter identification and a novel control method is then further given using the Lyapunov functional method. With this new and effective method, parameter identification and lag synchronization can be achieved simultaneously. Simulation results are given to justify the theoretical analysis in this paper.     

光学常数测量方法的探讨

本文讨论了运用宽光谱光电折射仪和Ｃ２０型精密测角仪，采用不同的方法对各种类型的光学材料进行了从紫外到近红外光谱区域内光学常数的测量，并着重介绍如何运用测角仪在可见光区域内对各种楔型块进行光学常数的测定     

基于Khatri-Rao子空间和传播算子的宽带声源波达方向估计算法

本文提出基于Khatri-Rao子空间和传播算子的宽带声源波达方向估计算法。该算法将声源不同频率处的协方差矩阵变换重排为一个高维矩阵,然后利用传播算子方法估计宽带声源波达方向。该算法计算复杂度介于聚焦Khatri-Rao子空间和相干子空间算法之间。仿真和实验结果表明,该算法在降低计算量的同时,估计误差与聚焦Khatri-Rao子空间算法相近,远小于相干子空间算法。     

基于信号子空间重构的鲁棒子区域Frost波束形成

本文针对一个感兴趣子区域内的宽带波束形成问题, 提出一种基于信号子空间重构的鲁棒Frost波束形成算法. 该算法的基本思想是利用一个矩阵滤波器从所估计的信号加干扰子空间中提取感兴趣信号(SOI)的特征分量; 然后利用该特征分量重构信号子空间; 最后利用重构的信号子空间构造一组线性约束最小方差(LCMV)准则, 来保证SOI近似无失真通过. 与现有的其他鲁棒Frost波束形成算法相比, 本文算法的一个显著优点是在未知SOI实际波达方向与频带等先验信息的情况下, 其导向与带宽均能够自适应地匹配SOI. 因此在整个感兴趣子区域内, 它都能获得接近最优值的输出信干噪比. 理论分析与仿真研究验证了算法的有效性.     

多子阵子空间拟合波达方向估计

提出了考虑阵元互耦的子空间拟合波达方向估计方法。用一带状、对称Toeplitz矩阵对均匀线阵的互耦矩阵建模。根据互耦矩阵自由度,去除部分两边阵元,研究了子空间拟合算法中ESPRIT算法的理论性能。在大快拍的情况下,均匀线阵的MCLS-ESPRIT算法的角度估计误差是一个零均值的联合高斯分布,同时给出了分布的协方差表达式。由方差的数值计算可知,在阵元数和信源数确定的情况下,存在估计标准差最小的子阵;与不存在互耦的理想情况相比,除了估计标准差少许增加外,性能接近。本文的方法已应用于高分辨率测深侧扫声呐,获得了好结果。在声呐阵阵元数增加的条件下,用本文的方法,能抑制均匀线阵的互耦的影响。