基于序列影像种子跟踪的道路线自动提取算法

为快速建立道路精确的几何模型,利用车载移动测量系统获取的序列影像开展了道路边线自动提取方法研究.针对不同类型道路边线的特征,提出了一种基于序列影像种子跟踪的道路线自动算法.首先,对不同质量道路影像分别采用基于灰度信息和梯度信息的道路边线边缘检测方法;其次,采用线性回归模型拟合道路边线,以当前帧拟合直线为种子点,提出了基于种子直线的序列影像感兴趣预测方法,在此基础上对下一帧影像道路边线进行跟踪;最后,以高速公路和城市街区道路序列影像为例,开展了道路边线自动跟踪提取试验,结果表明此方法可以达到优于20 cm的精度,具有很好的鲁棒性.     

HSV特异性转移因子治疗复发型HSVK的临床观察

利用单纯疱疹病毒特异性转移因子联合无环鸟苷滴眼液治疗４０例复发型单纯疮疹病毒性角膜炎（Ｈｅｒｐｅｓ　ｓｉｍｐｌｅｘ　ｖｉｒｕｓ　ｋｅｒａｔｉｔｉｓ，ＨＳＶＫ）患者，总有效率为１００％，治愈率为８６．６％。采用红细胞Ｃ３ｂ受体花环试验研究红细胞免疫功能的变化。其结果，１０例正常人与４０例治疗前患者比较（Ｐ＜０．０１），两组有显著性差异。４０例患者，治疗前与治疗后比较两组有显著差异（Ｐ＜０．０１）。表明ＨＳＶＫ患者治疗前免疫功能低下，用ＨＳＶ特异性转移因子治疗后，红细胞免疫功能显著提高。     

基于滞回分配的VBR视频流量带宽管理方案

基于VBR视频流量预测提出了一种新的动态带宽分配算法———滞回算法。与Girish提出的门限算法和Youssef提出的基于GOP场景检测的带宽分配方案相比,对缓存尺寸的需求、信元丢失率等性能有明显提高。大量的仿真表明,该算法能够以较低的开销(带宽重分配次数百分比≤2%)同时兼顾了CLR和带宽利用率,而且只有在缓存队长超过迂回门限时才触发带宽重分配过程,有利于在线实现。     

人脑胶质瘤nm23-H1蛋白及PCNA的表达

目的　研究人脑胶质瘤中nm2 3 H1蛋白和增殖细胞核抗原 (PCNA)表达状况。方法　采用免疫组化染色技术 ,对 5 3例人脑胶质瘤石蜡切片中nm2 3 H1蛋白和PCNA进行检测。结果　 5 3例人脑胶质瘤免疫组化染色结果显示 ,人脑胶质瘤I级、II级中nm2 3 H1阳性细胞显著高于Ⅲ、Ⅳ级 (P <0 .0 5 ) ;Ⅲ、Ⅳ级中PCNA阳性细胞显著高于I、II级(P <0 .0 5 ) ,但PCNA蛋白表达与nm2 3 H1表达无显著相关性。结论　nm2 3 H1蛋白低表达及PCNA高表达与脑胶质瘤的病理学分级相关 ;nm2 3 H1表达可能成为脑胶质瘤的浸润与转移的指标。     

低硒对心室肌纤维细胞动作电位的影响

用克山病区的粮食饲养大鼠,复制低硒动物模型、用浮置玻璃微电极技术,研究低硒大鼠在体心室肌细胞动作电位的变化,并与对照组进行了对比观察。结果发现低硒大鼠动作电位振幅明显降低(P<0.05),动作电位除极时限显著延长(P<0.01),而动作电位复极50%及90%的持续时间变化不显著(P>0.05)。实验结果说明低硒可明显影响大鼠心室肌细胞动作电位变化。为克山病心肌损害的病因学研究提供了重要的实验依据。     

多抗乙素对小鼠S_(180)肉瘤中免疫活性细胞浸润的影响

曾经证明多抗乙素有明显的抑瘤作用。本文以小鼠实验性S_(180)肉瘤为模型观察了多抗乙素对肿瘤组织内和肿瘤周围Mφ、L_3T_4~- 及Lyt_2~-细胞的影响,同时观察了肿瘤坏死组织周围中性粒细胞浸润及小血管内血栓形成的改变。结果提示:Pb治疗组,肿瘤周围及肿瘤内浸润的L_3T_4~-和Lyt_2~-细胞显著增多,说明Pb可诱导肿瘤组织中淋巴细胞浸润;肿瘤组织的坏死灶周围明显充血、小血管血栓形成显著增加,并有带状中性粒细胞浸润。这些改变与TNF介导的肿瘤组织损伤相符,提示Pb的抑瘤作用主要是通过免疫活性细胞产生的TNF实现的。     

原发性肾病综合征患儿T细胞亚群及其抑制性细胞活性

本文检测了37例原发性肾病综合征患儿的外周血T淋巴细胞亚群和刀豆素A诱导的抑制细胞活性水平。结果显示,肾病患儿初发时就存在抑制性T细胞数量及功能的减低,并随着激素治疗及病情的好转而恢复正常,提示本病的免疫调节失衡可能是一原发性改变,在肾病发病机理中起一定的作用。     

红细胞免疫功能检测—红细胞酵母花环法的改进

作者等对郭峰红细胞C_3b受体花环试验和红细胞免疫复和物花环试验的实验方法进行改进,建立了微量实验法。这一方法设计合理、步骤简便,节约试剂、方便病人,具有微量、准确、快速、简便、实用等优点,适宜于各级医疗单位推广应用。     

Impact of phytoplankton community size on a linked global ocean optical and ecosystem model

We isolated the effect phytoplankton cell size has on varying remote sensing reflectance spectra (R_rs(λ)) in the presence of optically active constituents by using optical and radiative transfer models linked in an offline diagnostic calculation to a global biogeochemical/ecosystem/circulation model with explicit phytoplankton size classes. Two case studies were carried out, each with several scenarios to isolate the effects of chlorophyll concentration, phytoplankton cell size, and size-varying phytoplankton absorption on R_rs(λ). The goal of the study was to determine the relative contribution of phytoplankton cell size and chlorophyll to overall R_rs(λ) and to understand where a standard band ratio algorithm (OC4) may under/overestimate chlorophyll due to R_rs(λ) being significantly affected by phytoplankton size. Phytoplankton cell size was found to contribute secondarily to R_rs(λ) variability and to amplify or dampen the seasonal cycle in R_rs(λ), driven by chlorophyll. Size and chlorophyll were found to change in phase at low to mid-latitudes, but were anti-correlated or poorly correlated at high latitudes. Phytoplankton size effects increased model calculated R_rs(443) in the subtropical ocean during local spring through early fall months in both hemispheres and decreased R_rs(443) in the Northern Hemisphere high latitude regions during local summer to fall months. This study attempts to tease apart when/where variability about the OC4 relationship may be associated with cell size variability. The OC4 algorithm may underestimate [Chl] when the fraction of microplankton is elevated, which occurs in the model simulations during local spring/summer months at high latitudes in both hemispheres.     

Optimal development of alternative fuel station networks considering node capacity restrictions

A potential solution to reduce greenhouse gas (GHG) emissions in the transport sector is the use of alternative fuel vehicles (AFV). As global GHG emission standards have been in place for passenger cars for several years, infrastructure modelling for new AFV is an established topic. However, as the regulatory focus shifts towards heavy-duty vehicles (HDV), the market diffusion of AFV-HDV will increase as will planning the relevant AFV infrastructure for HDV. Existing modelling approaches need to be adapted, because the energy demand per individual refill increases significantly for HDV and there are regulatory as well as technical limitations for alternative fuel station (AFS) capacities at the same time. While the current research takes capacity restrictions for single stations into account, capacity limits for locations (i.e. nodes) – the places where refuelling stations are built such as highway entries, exits or intersections – are not yet considered. We extend existing models in this respect and introduce an optimal development for AFS considering (station) location capacity restrictions. The proposed method is applied to a case study of a potential fuel cell heavy-duty vehicle AFS network. We find that the location capacity limit has a major impact on the number of stations required, station utilization and station portfolio variety.