广义系统多传感器分布式融合降阶Kalman滤波器

对于带多传感器的广义线性离散随机系统,应用奇异值分解,将其变换为等价的两个降阶多传感器子系统,提出了基于变换后的状态融合器构造原始状态融合器的新的融合方法。应用Kalman滤波方法,在线性最小方差按矩阵加权、按对角阵加权和按标量加权融合准则下,分别提出了三种最优加权融合降阶广义Kalman滤波器。可统一处理融合滤波、平滑和预报问题。可减少计算负担和改善局部滤波精度。证明了三种融合器和局部估值器之间的精度关系。为了计算最优加权。提出了局部滤波误差协方差阵的计算公式。一个Monte Carlo仿真例子说明了其有效性。     

基于小波能量加权的医学图像融合新算法

目前的大多数基于小波变换的医学图像融合算法,由于没有细致考虑低频分量融合规则以及高频分量邻域特征对于融合效果的重要性,因而得到的融合效果有时并不理想.现提出一种小波域按邻域加窗能量加权的融合准则,该准则根据人体颅部18F-FDG和MRI-T1图像的特点,充分考虑到了低频分量和高频分量邻域特征,对原图像高、低频小波系数分别选用具有不同物理意义的窗矩阵进行邻域加窗能量计算,以归一化加窗能量为权值对各小波系数进行加权,得到融合小波系数,经过小波反变换重构出融合图像.实验证明,采用提出的融合算法得到的融合图像很好地包含了源图像的信息,该算法在医学图像融合应用中比传统的小波系数取大算法、加权平均算法及按邻域方差加权算法效果更好.     

X射线荧光光谱法测定海底钴结壳的方法研究

采用粉末压片制样,X-射线荧光光谱法直接测定钴结壳中Fe、Mn、Co、Ni、Si、Al、Ca、K、Na、Ti、P、Mg、Cu、Pb、Mo、Sr、V、Zn、Zr、Y、Ba等元素。重点讨论用熔融法定值部分样品,然后用加以定值的钴结壳标准及样品、国家岩石、土壤、水系沉积物系列标准用粉末压片制样,做工作曲线,一次测量钴结壳中二十余种元素,并用定值样品做精密度与准确度试验,各元素统计结果为RSD％（n=12）除V的RSD〈4％外,其余元素的RSD均小于2％,测定结果与标准值和ICP-MS测量值相符。     

基于KULLBACK信息融合方法的系统可靠性评估

在产品具有多源验前信息的情形下,讨论冷贮备系统的可靠性评估问题,运用Kullback信息作为分布之间距离的度量,在Kullback信息的融合准则下对多个先验分布进行融合。并以融合后的先验分布作为产品的最终验前分布,对冷贮备系统的可靠性指标进行Bayes估计。最后进行的计算机随机模拟结果表明,文中所提出的方法合理且便于应用。     

多极化多特征融合的雷达目标识别研究

雷达多极化组合态距离像与单极化组合态距离像相比包含了更多的目标信息,因此也具有更强的目标鉴别能力;而距离像不同的特征同样包含了具有一定差异的鉴别信息,可以用来提高距离像的总体识别能力。基于D-S证据理论将多极化信息和六种常用的距离像特征信息有效融合起来,获得了较高的识别率,基于仿真数据的实验验证了结论的正确性。     

基于NSCT变换的红外与可见光图像融合算法

分析了非子采样Contourlet变换滤波器组的设计与实现方法,提出一种基于非子采样Contourlet变换的红外与可见光图像融合算法。首先将图像作非子采样拉普拉斯金字塔尺度分解,并在各个尺度层对高频子带作非子采样方向分解;然后分别采用基于区域能量和区域方差的的融合规则得到融合图像的非子采样Contourlet低频系数和高频系数;最后再进行非子采样Contourlet逆变换得到融合图像。实验结果表明,该方法的融合效果优于àtrous小波变换方法和Contourlet变换方法。     

基于Bayes估计的多传感器数据融合方法研究

对多传感器数据融合方法进行研究 ,以 Bayes估计理论为基础得到了多传感器最优融合数据 ,并将它与其它方法得到的融合数据进行了比较 .     

<Emphasis Type="Italic">Legionella pneumophila</Emphasis> — a human pathogen that co-evolved with fresh water protozoa

The bacterial pathogen Legionella pneumophila is found ubiquitously in fresh water environments where it replicates within protozoan hosts. When inhaled by humans it can replicate within alveolar macrophages and cause a severe pneumonia, Legionnaires disease. Yet much needs to be learned regarding the mechanisms that allow Legionella to modulate host functions to its advantage and the regulatory network governing its intracellular life cycle. The establishment and publication of the complete genome sequences of three clinical L. pneumophila isolates paved the way for major breakthroughs in understanding the biology of L. pneumophila. Based on sequence analysis many new putative virulence factors have been identified foremost among them eukaryotic-like proteins that may be implicated in many different steps of the Legionella life cycle. This review summarizes what is currently known about regulation of the Legionella life cycle and gives insight in the Legionella-specific features as deduced from genome analysis. Received 1 September 2006; received after revision 10 October 2006; accepted 22 November 2006     

Targeted inhibition of Livin resensitizes renal cancer cells towards apoptosis

Cancer cells are typically characterized by apoptosis deficiency. In order to investigate a possible role for the anti-apoptotic livin gene in renal cell cancer (RCC), we analyzed its expression in tumor tissue samples and in RCC-derived cell lines. In addition, we studied the contribution of livin to the apoptotic resistance of RCC cells by RNA interference (RNAi). Livin gene expression was detected in a significant portion of RCC tumor tissue specimens (13/14, 92.9%) and tumor-derived cell lines (12/15, 80.0%). Moreover, targeted inhibition of livin by RNAi markedly sensitized RCC cells towards proapoptotic stimuli, such as UV irradiation or the chemotherapeutic drugs etoposide, 5-fluorouracil, and vinblastine. These effects were specific for livin expressing tumor cells. We conclude that livin can contribute significantly to the apoptosis resistance of RCC cells. Targeted inhibition of livin could represent a novel therapeutic strategy to increase the sensitivity of renal cancers towards pro-apoptotic agents. Received 30 November 2006; received after revision 22 February 2007; accepted 20 March 2007     

基于启发式的多机器人SLAM地图融合方法研究

同步定位和地图构建（simultaneous localization and mapping,SLAM）是移动机器人在未知环境中完成地图构建和定位任务的关键技术。针对多机器人SLAM中的地图融合问题,提出一种启发式的搜索方法引导局部地图的重复区域进行地图融合。每个机器人可以在不了解其相对位置的情况下建立局部地图,并将局部地图信息发送至同一工作站中,以局部地图的相似性为判断指标融合得到最优的全局地图。在机器人实物平台上进行验证,结果证明了多机器人SLAM的地图融合算法的有效性和准确性。