Land cover classification of the North China Plain using MODIS_EVI time series

Moderate Resolution Imaging Spectroradiometer (MODIS) data have played an important role in global environmental and resource research. However, its low spatial resolution has been an impediment to researchers pursuing more accurate classification results. In this research, the high temporal resolution of MODIS was employed to improve the accuracy of land cover classification of the North China Plain using MODIS_EVI time series from 2003. Harmonic Analysis of Time Series (HANTS) was performed on the MODIS_EVI image time series to reduce cloud and other noise effects. The improved MODIS_EVI time series was then classified into 100 clusters by the Iterative Self-Organizing Data Analysis Technique (ISODATA). To distinguish ambiguous land cover classes, a decision tree was built on five phenological features derived from EVI profiles, Land Surface Temperature (LST) and topographic slope. The overall accuracy of the final land cover map was 75.5%, indicating the promise of using MODIS EVI time series and decision trees for broad area land cover classification.     

基于物联网三层架构的地下工程测量机器人远程变形监测系统

针对地下工程变形监测特点,在物联网三层架构技术和测量机器人技术的基础上,以C#编程语言和SQL sever数据库为开发平台,借助GeoCOM开发包,通过远程控制测量机器人对地下工程进行自动化变形监测工作。在物联网三层技术架构自动化变形监测系统应用服务层中,介绍了提高数据处理精度的实时差分技术原理和围岩净空收敛分析方法。成功地将物联网技术应用在变形监测领域,保证了地下工程的安全服务;而这一系统设计思想更可为同类工程安全监测提供参考。     

一种基于SVM的无人机影像中单个建筑物的角点检测方法

针对目前无人机影像中单个建筑物角点的检测现状,提出了一种基于支持向量机(SVM)的无人机影像中建筑物的角点检测方法。首先对4个波段的无人机影像进行多尺度分割,计算影像的NDVI,通过植被与非植被区域的波谱差异剔除植被的影响。其次,用面向对象分类法将"建筑物块"从影像中提取出来,对"建筑物块"区域用Harris算子进行边缘检测,形成建筑物边缘点集数据。随后通过设计高斯径向基将边缘样本点映射到高维特征空间,构建特征向量,采用边缘点集训练SVM分类模型,最终通过SVM分类模型从粗提取的边缘点集中检测出正确的建筑物角点,实现了单个建筑物的角点提取。     

无人机自主着陆图像识别压缩技术研究

近些年来,随着遥感测绘需求日益增加,测绘型旋翼无人机的使用变得愈加频繁。测绘型旋翼无人机以其独特的气动性能及应用特点,使得其与固定翼无人机可以共同应用于航片拍摄。目前,由于地形及地域等多重因素的影响,测绘型旋翼无人机在执行任务期间经常遇到气流扰动、遥控信号干扰、GPS辅助信号中断等意外情况,需要应用其自主模式进行辅助着陆。而要实现自主着陆首先需要解决的是无人机对地面着陆地标的识别处理等一系列问题,本文就相关的图像识别压缩技术进行设计研究。     

Improved GNSS-based indoor positioning algorithm for mobile devices

GPS Solutions - The widespread use of global navigation satellite system (GNSS) receiver in mobile devices induces the adoption of effective GNSS-based indoor positioning algorithms exploiting...     

Assessment of agricultural drought in rainfed cereal production areas of northern China

Agricultural drought assessment is an important tool for water management in water-scarce regions such as Inner Mongolia and northeastern China. Conventional methods have difficulty of clarifying long-term influences of drought on regional agricultural production. To accurately evaluate regional agricultural drought, we assessed the performance of drought indices by constructing a new assessment framework with three components: crop model calibration and validation, drought index calculation, and index assessment (standard period setting, mean value and agreement assessments). The Environmental Policy Integrated Climate (EPIC) model simulated well of county-level wheat and maize yields in the nine investigated counties. We calculated a major crop-specific index yield reduction caused by water stress (WSYR) in the EPIC crop model, by relating potential and rainfed yields. Using 26 agricultural drought cases, we compared WSYR with two meteorological drought indices: precipitation (P) and aridity index (AI). The results showed that WSYR had greater agreement (85 %) than either the precipitation (65 %) or aridity index (68 %). The temporal trend of the indices over the period 1962–2010 was tested using three approaches. The result via WSYR revealed a significant increase in the trend of agricultural drought in drought-prone counties, which could not be shown by the precipitation and aridity indices. Total number of dry year via WSYR from 1990s to 2000s increases more sharply than via P or AI. As shown by WSYR, the number of dry years in northeastern China and Inner Mongolia is generally increasing, particularly after the 2000s, in the western part of the study area. The study reveals the usefulness of the framework for drought index assessment and indicates the potential of WSYR and possible drought cases for drought classification.     

川芎嗪对铁过载原代培养脐静脉内皮细胞超氧化物歧化酶、活性氧表达的影响

目的 探讨铁过载对原代培养脐静脉内皮细胞（HUVECS）超氧化物歧化酶（SOD）、活性氧表达的影响及川芎嗪的干预效应。方法 使用酶联免疫法（ELISA）检测体外培养铁过载脐静脉内皮细胞表达的超氧化物歧化酶和活性氧的含量。结果 与空白组比较,模型组超氧化物歧化酶的含量显著降低（P<0.01）,给药组超氧化物歧化酶与模型组比较,差异有统计学意义（P<0.05）;模型组活性氧含量显著高于空白组（P<0.01）,与模型组相比,给药组活性氧含量有明显降低（P<0.05）。结论 川芎嗪能抑制脐静脉内皮细胞铁过载导致的氧化应激,有通过抗氧化应激发挥抗动脉粥样硬化的作用。     

黄河流域植被覆盖度动态变化与降水的关系

利用8km分辨率Pathfinder NOAA-NDVI数据，对黄河流域1982－1999年地表植被覆盖的空间分布及时间序列变化进行了分析，并通过计算不同时段降水量与年最大NDVI之间的相关系数分析了降水对流域植被覆盖的影响。结果发现近20年来黄河流域平均植被覆盖度有增加趋势，但青藏高原上有所减小；汛期降水量的多少对地表植被覆盖度的年际变化起主要作用，其中草原地区影响最显著，而在森林植被区及部分灌溉农作区，降水的年际变化对地表覆盖的影响比较小。     

深地震测深数据子库

基于Microsoft Access97数据库开发工具和支持ActiveX,OLE与自动化对象的通用语言VBA开发的以查询，调用，处理和共享为目的的深地震测深数据库管理系统和基于国土资源部国际合格与科技司下发的深地震测深专业方法数据入库技术规定，收集，抢救了原地质矿产部“九五”以前完成的14条宽带深地震测深剖面的原始记录和文字资料，通过编辑，整理，统一地震道集记录存取格式与入库，建立了原始数字地震道集记录，成果目录与图形数据管理，维护与预处理平台，实现了让深地震测深数据活起来，达到能再次开发利用的科研目标。