基于BP神经网络的人脸识别

将BP神经网络用于人脸识别 ,建立了人脸识别模型 ,研究了样本采样训练、样本批量训练和样本完整训练三种训练策略对识别率的影响。所设计的识别模型包括图像压缩、图像抽样、输入矢量标准化、BP神经网络与竞争选择处理过程。利用ORL人脸图像数据库进行了仿真实验 ,结果表明 ,其识别模型在实际应用中是可行的。该模型简单 ,识别率较高。如将训练策略配合使用 ,则在提高训练速度和训练效率的同时 ,也使模型分类性能有了明显提高     

电子电路设计中异构印刷电路板数据融合技术

论述了当前电子设计自动化 (electronicdesignautomatic ,EDA)设计中普遍存在的异地异构设计数据互不兼容而产生的一系列问题。在分析了市场需求的基础上 ,引出了EDA协同设计集成环境的概念 ,并以Cadence和Mentor的印刷电路板 (printedcircuitboard ,PCB)设计数据融合技术为例进行了详细阐述 ,为EDA工具在PCB级的集成提出了一条崭新的思路。     

A Fuzzy Neural Model for Face Recognition

Introduction1 Nowadays, biometric (e.g., face, finger print, iris, palm)identification is becoming more and more important [1] [2] and there are numerous commercial and law enforcement applications. Compared with other biometric identifications, face recognition has attracted more attention for its characteristics ofuser-friendly, intuitionistic and convenient. Much work hasbeen done on face recognition over the last 20 years. Generally speaking, face recognition is composed of three procedure…     

一种基于弱标签的三维模型语义自动标注方法

随着三维应用的普及,三维模型大量产生并广泛传播。由于三维模型广泛应用于计算机辅助设计、三维游戏、电影特效制作等诸多领域,已经形成了大量的三维模型数据库。三维模型语义标注的目的是给出描述其语义的标注词,是三维模型管理和基于文本的三维检索的关键技术。针对互联网大量存在的弱标签三维模型现状,提出一种基于弱标签的三维模型语义标注方法LPMLL,首先,采用半监督学习方法进行标签传播,得到标注词置信度,达到提升训练集的目的。然后,采用一种基于最大后验概率准则的方法进行多标签学习,得到最终标注词。实验数据表明了该方法的有效性。     

基因表达式编程在组合预测建模中的应用

介绍一种利用基因表达式编程的方法来自动生成非线性函数的组合预测模型,并进行误差估计分析, 改变过去只依靠各个子方法的简单线性相加,不能很好地反映非线性真实世界的传统组合预测建模方法.通过对我国CPI的真实历史数据验证, 验证结果表明: 与传统的ARIMA,灰色GM(1,1), BP神经网络和线性组合预测四种方法对比,基因表达式编程建立的组合预测模型所预测的数据准确度明显提高.     

非常规突发事件中面向目标能力的路网调整及车流组织模型

在突发事件的背景下, 事件影响区域的交通路网是重要的生命线, 能否基于交通管制手段优化路网结构, 有效协调救援、疏散等任务于路网上的时空干扰, 是决定应急响应成功与否的重要因素. 首先从宏观上设计了该背景下的多层加载式的路网调整框架; 然后, 基于元胞传输模型(CTM)分析了各层交通需求加载时路网调整的目标函数及相应的系统约束、交通特性约束和优先级约束等, 最终构成了系统的路网调整模型; 最后, 通过算例对模型进行了验证分析.     

ISAR target recognition based on non-negative sparse coding

Aiming at technical difficulties in feature extraction for the inverse synthetic aperture radar (ISAR) target recognition, this paper imports the concept of visual perception and presents a novel method, which is based on the combination of non-negative sparse coding (NNSC) and linear discrimination optimization, to recognize targets in ISAR images. This method implements NNSC on the matrix constituted by the intensities of pixels in ISAR images for training, to obtain non-negative sparse bases which characterize sparse distribution of strong scattering centers. Then this paper chooses sparse bases via optimization criteria and calculates the corresponding non-negative sparse codes of both training and test images as the feature vectors, which are input into k neighbors classifier to realize recognition finally. The feasibility and robustness of the proposed method are proved by comparing with the template matching, principle component analysis (PCA) and non-negative matrix factorization (NMF) via simulations.     

基于边界矩和改进FCM聚类的水下目标识别

水下目标的识别是水下机器人对环境动态感知、快速定位与跟踪视觉目标的关键, 本文针对水下成像的特殊性以及成像环境的复杂性,旨在设计一种快速、准确的目标识别系统以指导水下机器人进行下一步的任务. 首先, 综合运用一些流行的算法, 简要介绍了一种有效的边界分割算法; 然后通过对边界矩的分析和修正, 构造了具有平移、旋转及比例变换不变性的仿射变换; 最后详细描述了改进的FCM聚类识别的设计理念. 通过对实测的4类物体组成的水下目标的识别实验, 证明了所提水下目标识别系统可以用于水下目标识别, 并且具有较高的鲁棒性和实时性.     

Energy efficient target tracking algorithm using cooperative sensors

Target tracking is one of the applications of wireless sensor networks(WSNs).It is assumed that each sensor has a limited range for detecting the presence of the object,and the network is sufficiently dense so that the sensors can cover the area of interest.Due to the limited battery resources of sensors,there is a tradeoff between the energy consumption and tracking accuracy.To solve this problem,this paper proposes an energy efficient tracking algorithm.Based on the cooperation of dispatchers,sensors in the area are scheduled to switch their working mode to track the target.Since energy consumed in active mode is higher than that in monitoring or sleeping mode,for each sampling interval,a minimum set of sensors is woken up based on the select mechanism.Meanwhile,other sensors keep in sleeping mode.Performance analysis and simulation results show that the proposed algorithm provides a better performance than other existing approaches.     

Target estimation using MIMO radar with multiple subcarriers

This paper analyzes the effect of waveform parameters on the joint target location and velocity estimation by a noncoherent multiple input multiple output (MIMO) radar transmitting multiple subcarriers signals. How the number of subcarriers influences the estimation accuracy is illustrated by considering the joint Cramer-Rao bound and the mean square error of the maximum likelihood estimate. The non-coherent MIMO radar ambiguity function with multiple subcarriers is developed and investigated by changing the number of subcarriers, the pulse width and the frequency spacing between adjacent subcarriers. The numerical results show that more subcarriers mean more accurate estimates, higher localization resolution, and larger pulse width results in a worse performance of target location estimation, while the frequency spacing affects target location estimation little.