基于无人机传感器数据与词袋模型的点云自动化分类

激光点云分类是Li DAR数据应用的关键步骤和重要研究课题。针对Li DAR点云数据识别率低的问题,以体素化的点云为研究对象,提出了一种基于词袋模型的城区机载Li DAR点云数据分类方法。考虑到点云数据缺乏纹理信息,文中综合分析了点云数据和影像数据的特点,以点为单位提取描述点云的几何特征和影像特征分类特征;以体素为单位分割点云数据,并以体素为基础构建描述场景信息的词袋模型;最后基于随机分类器完成场景的分类。文中以ISPRS提供的Vaihingen数据作为实验数据。实验结果表明,本文提出的模型能有效地改善点云的分类质量,分类正确率能达到93%以上。     

近地机载激光雷达棉花LAI提取方法研究

棉花叶面积指数（LAI）是描述其长势的重要指标,准确获取冠层结构参数是叶面积指数反演的必要条件。以ScoutB-100油动单旋翼无人机为飞行平台,搭载RIEGL VUX-1激光雷达,精确获取棉花高密度点云数据,得到研究区棉田数字表面模型（DSM）和数字高程模型（DEM）,通过差值运算获得其冠层高度模型（CHM）,进而提取有效的冠层结构参数。利用相关性分析法选取相关系数大于0.2的激光穿透力指数（LPI）、回波点云密度（D）、孔隙率（f_gap）、归一化高程值（V_nDSM）构建棉花LAI反演模型,并与实测叶面积指数进行精度验证与评价。实验结果表明：模型估算的LAI与实测LAI之间的决定系数为0.824,均方根误差为0.072,验证了模型的可靠性。     

萍乐坳陷西部地区航测遥感油气预测研究

建立了萍乐坳陷西部地区的航测、遥感油气预测模型。通过航测、遥感构造解译和区域油气成藏背景分析,指出该地区具有中小型油气藏成藏的构造条件。圈出了常指示油气圈闭构造的航磁局部构造异常和遥感环形构造,并提取了与油气微渗漏直接相关的微磁异常、航空放射性低铀异常、低钾异常和热异常,据此预测出三片油气成藏有利区。     

Study on Size Distributions of Airborne Particles by Aircraft Observation in Spring over Eastern Coastal Areas of China

The authors studied the size distributions of particles at an altitude of 2000 m by aircraft observation over eastern costal areas of China from Zhuhai, Guangdong to Dalian, Liaoning (0.3,7-30μm, 57 channels,including number concentration distribution, surface area concentration distribution and mass concentration distribution). In these cities, the average daily concentrations of PM10 are very high. They are among the most heavily polluted cities in China. The main pollution sources are anthropogenic activities such as wood, coal and oil burning. The observed size distributions show a broad spectrum and unique multi-peak characteristics, indicating no significant impacts of individual sources from urban areas. These results are far different from the distribution type at ground level. It may reflect the comprehensive effect of the regional pollution characteristics. Monitoring results over big cities could to some extent reflect their pollution characteristics.     

The simulation of airborne disease virus transmission in group buildings

With the densification of population and the deterioration of urban ecological cycle, the airborne disease virus is becoming a new crisis for human. It is well known that airborne disease virus such as smallpox, hives, mumps and SARS can cause respiratory disease by the aerosol of sneezes or cough of infected person. The pathogens release virus aerosol to the environment. If the virus can live some time in the air, it can go with air flowing far away. The aim of the present research is to simulate the transportation of airborne disease virus in a group of buildings by taking into account the time-dependent virus aerosol source. A commonly used virus-host dynamical model is introduced to simulate the time-dependent development of virus aerosol releasing process. The wind velocity field is described by the continuity equation, momentum equation and turbulence model to enclose the Reynolds stress in the momentum equation. The RNG model is used in all simulations. The aerosol of microbial pollutant is assumed flowing at the same velocity as the air. The computation time is one life cycle of virus T. We assume that the pathogens release virus aerosol at the upper wall of building A. The conclusions are： （1） The microbial pollutant begins to be discharged after its delitescence 1/12 T. As the virus goes into logarithmic growth period, the concentration of virus and the pathogenic area increase very fast. After entering the virus decay period （1/3 T）, the pathogenic area decreases. It should be noticed that no secondary infection is considered in present simulation. Generally, as the virus concentration is above the pathogenic concentration, new pathogens will be infected and release virus aerosol to the environment. With the increase of new pathogens, the pathogenic area would increase very fast and finally result in the outbreak of infection.     

北方中新生代产铀盆地的特点与找矿方向探讨

根据北方主要中新生代盆地的沉积演化特征,结合北方地区已知产铀盆地的特点,通过对航空物探测量成果资料的分析,建立了找矿选区依据;通过初步分析,指出了北方主要中新生代盆地砂岩型铀矿的找矿方向。     

机载InSAR数据自动生成DEM技术及其在内蒙古丰镇地区的应用

合成孔径雷达干涉测量(InSAR)技术是传统的合成孔径雷达技术和射电天文学中的干涉测量技术相结合而发展起来的一项新的遥感技术,这一新技术以其在大范围地表高程测量和地表变形测量方面所具有的独特优势和巨大潜力,而成为对地测量和地学科学研究的一个重要的工具。同星载InSAR相比,机载InSAR在高分辨率区域测图方面具有较大的优势,而且数据采集的时间安排及飞行方位选取方面相对较灵活。而且SAR是一种主动微波遥感技术,具有全天时全天候的工作特征,这一技术也正在成为中国西部困难地区测图的一个非常有效的工具。在详细分析应用机载InSAR数据测量地表高程的基本原理的基础上,提出了机载InSAR自动生成DEM的技术流程,并以内蒙古丰镇地区的机载InSAR数据为基础进行了试验研究,取得了较好的效果,为这一新技术的进一步实际应用奠定了基础。     

L＆R航空重力仪的水平加速度改正

L&R(LaCoste&Romberg)航空重力仪是我国首套航空重力测量系统的核心,主要由高精度的垂直加速度计和稳定平台组成.前者用于测量总加速度,后者使加速度计保持精确的垂直指向.飞行测量时,由于水平加速度的存在,稳定平台难以精确维持水平从而使加速度计偏离正确指向,由此产生了水平加速度改正.介绍了水平加速度计算的两种基本方法,即两步法与一步法,从理论上导出了其在平台倾角较小情况下的等价性.为有效地减弱因水平加速度改正不完善产生的系统性误差,提出了水平加速度改正的预滤波方法.通过平台倾角的谱分析,确定了预滤波尺度,即采用与平台稳定周期相当的低通滤波器进行预滤波.最后利用实测数据对此进行了验证和分析.结果表明, 利用预滤波尺度, 航空重力测量值与地面向上延拓参考值的系统差由5×10-5 m·s-2减小至0.5×10-5 m·s-2.