基于正交多项式的平原城区机载LiDAR数据滤波算法

从机载LiDAR点云数据中分离地面点与非地面点生成城区DEM,是构建数字城市的首要工作。该文对机载LiDAR数据处理中的滤波关键算法进行了研究。首先,利用常用的区域增长方法对机载LiDAR点云数据进行了滤波处理,然后基于正交多项式分带滤波方法进行了点云滤波,比较了两种算法的运行效率,分析了二者的特点,并通过试验对比进行了阈值参数优选,试验表明该算法具有较好的滤波效果,在平原城区的数字城市三维建模中具有很好的应用价值。     

基于三角形驱动的机载LiDAR数据栅格化算法

针对基于TIN的机载LiDAR点云数据插值为GRID过程中,大量TIN数据I/O操作导致的栅格化效率下降问题,提出了基于三角形驱动的点云栅格化流式算法:以基于直线正负区判别原理的TIN向GRID转换新算法为基础,通过遍历三角形模拟流计算实现三角形生成、三角形插值为GRID以及内存释放,有效避免了TIN数据的I/O操作,并可以提高内存利用率。试验表明:流式算法显著提高了点云栅格化效率,为海量点云的栅格化及并行计算提供了一种算法支撑。     

基于ENVISAT ASAR数据的森林蓄积量估测研究

以东北大兴安岭地区的图强林业局作为试验区,基于统计方法对多时相ENVISAT ASAR数据进行森林蓄积量估测研究,分别选取HH、HV及HV/HH值分析森林蓄积量与SAR后向散射系数之间的关系;根据其散点图形状进行方程拟合及优选,完成蓄积量的估测。利用独立的检验样本对估测结果进行评价,均方根误差(RMSE)为16.526 m3/hm2,与实际情况基本相符。     

基于LiDAR数据的流域水系网络提取方法研究

LiDAR技术能够提供高分辨率DEM数据,可用于准确提取流域水系网络信息。研究中针对LiDAR系统提供的高精度DEM数据中存在的噪声,提出采用各向异性扩散滤波算法进行噪声平滑,与常用的DEM滤波算法比较,该方法既能有效去除噪声,同时又能保留高梯度的地形信息。在DEM噪声去除的基础上,采用基于局部地形曲面几何分析的Peuker&Douglas算法初步提取水系网络原型,进一步利用改进的基于坡面流物理模拟分析的加权D8算法提取水系网络,构建了基于LiDAR数据水系网络提取的技术流程。通过鹤壁市某小流域的LiDAR数据水系网络提取试验证明了该方法准确提取水系网络的有效性。     

基于动态椭圆的LiDAR数据窗户边界提取研究

针对复杂建筑物立面中窗户精细提取的难题,该文基于车载LiDAR数据提出一套完整的不规则窗户边界提取方法:先通过RANSAC算法探测建筑物主墙面点云,借助语义特征分离墙面和非墙面点云;再通过坐标变换,将非墙面点云转换到水平面,采用格网对水平面内点集的邻域关系进行判断,使用平面八邻域连通性探测方法对二维平面内窗户点进行聚类,分别存储聚类后的每一个窗户点云;然后采用改进的动态椭圆凸壳算法,检测聚类后各窗户边界轮廓点;最后对窗户点云坐标逆变换,得到建筑物立面中的窗户边界。通过实验验证了该算法的准确性。     

一种基于机载LiDAR和离散曲率的建筑物三维重建方法

机载LiDAR作为一种获取地物空间信息的新技术已得到广泛应用,但从LiDAR数据中重建建筑物三维模型方法的缺乏,是制约其进一步发展的瓶颈.该文介绍了一种自动与人机交互建模相结合的建筑物三维重建方法,首先基于离散曲率分析自动提取建筑物3D轮廓信息,然后将建筑物轮廓信息作为约束条件,使用模型驱动的ATOP算法实现建筑物自动建模,并支持对复杂建筑物人机交互建模.     

顾及矢量特征的机载LiDAR点云数据压缩方法

LiDAR技术可获取海量点云数据,数据压缩是点云数据预处理的一个重要环节。该文结合大范围城市区域高程范围随区域变化大的特点,提出一种有效的压缩策略。该策略主要通过建立高程方向不均等的三维格网来提高邻域点集的搜索效率,并根据某一压缩准则计算特征量对数据进行压缩。将该策略应用于上海浦东某区域,并与均匀采样法进行比较,证明其能有效提高点云数据压缩效率。     

基于航空影像和LiDAR数据的海地地震滑坡识别研究

快速、准确地判别滑坡灾害范围对灾情评估与救灾指挥至关重要。将高分辨率航空影像和机载LiDAR数据相结合,能充分利用滑坡体的光谱特征、几何形态与空间结构特征,提高滑坡识别的准确度。该文以海地地震诱发的滑坡识别为例,介绍了面向对象的滑坡识别思路及方法;针对海地滑坡灾害特点,选择了相应的灾害特征,并经反复试验设定了特征参数阈值;针对面向对象分类过程中特征组合与选择这一瓶颈问题,以河流沉积物和滑坡判别为例,通过特征组合试验与对比分析,发现基于NDVI与坡度特征组合的分类结果明显优于单一特征及其它组合模式的分类结果。与最大似然分类法相比,该方法的分类结果更为准确、精度更高。     

基于无人机成像光谱技术的农田土壤养分估测及制图

成像光谱技术是精准农业中土壤理化特性快速获取和实时监测的重要技术支撑.该文以松辽平原农田土壤为研究对象,通过对比标准正态变量变换(SNV)、多元散射校正(MSC)、光谱一阶微分(FD)和二阶微分(SD)4种方法对无人机高光谱图像进行预处理,引入竞争自适应重加权采样—连续投影组合算法(CARS-SPA)进行特征波段筛选,然后利用粒子群算法(PSO)对极限学习机算法(ELM)进行改进,从而构建土壤有机质(SOM)和土壤全氮(STN)的高光谱反演模型.结果表明:4种预处理方法中MSC去噪效果最优,可有效降低散射效应噪声;CARS方法筛选出53个SOM、33个STN特征波段,进一步利用CARS-SPA筛选出24个SOM和22个STN特征波段,将SOM和STN的平均相对误差分别降低了12.3％ 和6.6％;利用PSO优化ELM的输入权值和阈值,有效弥补了极限学习机模型泛化性能低这一不足,将SOM和STN的平均相对误差分别降低了2.9％ 和3.2％,同时模型精度高于偏最小二乘回归(PLSR)和支持向量机(SVM);利用CARS-SPA筛选的特征波段建立的PSO-ELM模型效果最佳,其预测SOM的决定系数R2为0.73、相对分析误差RPD为1.91,预测STN的决定系数R2为0.63、相对分析误差RPD为1.53,将模型应用于高光谱图像,获取SOM和STN空间分布图,玉米种植区的SOM和STN均高于花生种植区.该研究可为无人机高光谱图像应用于田间尺度的土壤养分估测和数字制图提供参考.     

基于机载极化雷达技术的农作物覆盖区土壤水分估算

从研究区辅助数据中获得植被覆盖区域农作物含水量，并对“水云模型”的参数进行了修改。从总的极化雷达后向散射中去除了植被影响；再应用简化的裸露地表模型和多极化数据特征消除了地表粗糙度的影响，从而能够通过多极化雷达数据得到地表土壤水分。经过文测数据验证表明农作物覆盖地表土壤水分变化的估算算法还需要更进一步发展和改进以提高反演精度，并探讨了应用极化雷达技术估算植被覆盖区土壤水分的方法。     

Association of elevation error with surface type,vegetation class and data origin in discrete-returns airborne LiDAR

Airborne LiDAR (light detection and ranging) data are now commonly regarded as the most accurate source of elevation data for medium-scale topographical modelling applications. However, quoted LiDAR elevation error may not necessarily represent the actual errors occurring across all surfaces, potentially impacting the reliability of derived predictions in Geographical Information Systems (GIS). The extent to which LiDAR elevation error varies in association with land cover, vegetation class and LiDAR data source is quantified relative to dual-frequency global positioning system survey data captured in a 400-ha area in Ireland, where four separate classes of LiDAR point data overlap. Quoted elevation errors are found to correspond closely with the minimum requirement recommended by the American Society of Photogrammetry and Remote Sensing for the definition of 95% error in urban areas only. Global elevation errors are found to be up to 5 times the quoted error, and errors within vegetation areas are found to be even larger, with errors in individual vegetation classes reaching up to 15 times the quoted error. Furthermore, a strong skew is noted in vegetated areas within all the LiDAR data sets tested, pushing errors in some cases to more than 25 times the quoted error. The skew observed suggests that an assumption of a normal error distribution is inappropriate in vegetated areas. The physical parameters that were found to affect elevation error most fundamentally were canopy depth, canopy density and granularity. Other factors observed to affect the degree to which actual errors deviate from quoted error included the primary use for which the data were acquired and the processing applied by data suppliers to meet these requirements.     

3D building roof reconstruction from airborne LiDAR point clouds: a framework based on a spatial database

Three-dimensional (3D) building models are essential for 3D Geographic Information Systems and play an important role in various urban management applications. Although several light detection and ranging (LiDAR) data-based reconstruction approaches have made significant advances toward the fully automatic generation of 3D building models, the process is still tedious and time-consuming, especially for massive point clouds. This paper introduces a new framework that utilizes a spatial database to achieve high performance via parallel computation for fully automatic 3D building roof reconstruction from airborne LiDAR data. The framework integrates data-driven and model-driven methods to produce building roof models of the primary structure with detailed features. The framework is composed of five major components: (1) a density-based clustering algorithm to segment individual buildings, (2) an improved boundary-tracing algorithm, (3) a hybrid method for segmenting planar patches that selects seed points in parameter space and grows the regions in spatial space, (4) a boundary regularization approach that considers outliers and (5) a method for reconstructing the topological and geometrical information of building roofs using the intersections of planar patches. The entire process is based on a spatial database, which has the following advantages: (a) managing and querying data efficiently, especially for millions of LiDAR points, (b) utilizing the spatial analysis functions provided by the system, reducing tedious and time-consuming computation, and (c) using parallel computing while reconstructing 3D building roof models, improving performance.     

近5年青海省植被覆盖变化的遥感监测

This paper used five years （2001-2006） time series of MODIS NDVI images with a 1-km spatial resolution to produce a land cover map of Qinghai Province in China. A classification approach for different land cover types with special emphasis on vegetation, especially on sparse vegetation, was developed which synthesized Decision Tree Classification, Supervised Classification and Unsupervised Classification. The spatial distribution and dynamic change of vegetation cover in Qinghai from 2001 to 2006 were analyzed based on the land cover classification map and five grade elevation belts derived from Qinghai DEM. The result shows that vegetation cover in Qinghai in recent five years has been some improved and the area of vegetation was increased from 370,047 km^2 in 2001 to 374,576 km^2 in 2006. Meanwhile, vegetation cover ratio was increased by 0.63%. Vegetation cover ratio in high mountain belt is the largest （67.92%） among the five grade elevation belts in Qinghai Province. The second largest vegetation cover ratio is in middle mountain belt （61.80%）. Next, in the order of the decreasing vegetation cover ratio, the remaining grades are extreme high mountain belt （38.98%）, low mountain belt （25.55%） and flat region belt （15.46%）. The area of middle density grassland in high mountain belt is the biggest （94,003 km^2）, and vegetation cover ratio of dense grassland in middle mountain belt is the highest （32.62%）, and the increased area of dense grassland in high mountain belt is the greatest （1280 km^2）. In recent five years the conversion from sparse grass to middle density grass in high mountain belt has been the largest vegetation cover variation and the converted area is 15931 km^2.     

Spatial and temporal variations of vegetation in Qinghai Province based on satellite data

This paper used five years (2001-2006) time series of MODIS NDVI images with a 1-km spatial resolution to produce a land cover map of Qinghai Province in China. A classi-fication approach for different land cover types with special emphasis on vegetation, espe-cially on sparse vegetation, was developed which synthesized Decision Tree Classification, Supervised Classification and Unsupervised Classification. The spatial distribution and dy-namic change of vegetation cover in Qinghai from 2001 to 2006 were analyzed based on the land cover classification map and five grade elevation belts derived from Qinghai DEM. The result shows that vegetation cover in Qinghai in recent five years has been some improved and the area of vegetation was increased from 370,047 km2 in 2001 to 374,576 km2 in 2006. Meanwhile, vegetation cover ratio was increased by 0.63%. Vegetation cover ratio in high mountain belt is the largest (67.92%) among the five grade elevation belts in Qinghai Prov-ince. The second largest vegetation cover ratio is in middle mountain belt (61.80%). Next, in the order of the decreasing vegetation cover ratio, the remaining grades are extreme high mountain belt (38.98%), low mountain belt (25.55%) and flat region belt (15.46%). The area of middle density grassland in high mountain belt is the biggest (94,003 km2), and vegetation cover ratio of dense grassland in middle mountain belt is the highest (32.62%), and the in-creased area of dense grassland in high mountain belt is the greatest (1280 km2). In recent five years the conversion from sparse grass to middle density grass in high mountain belt has been the largest vegetation cover variation and the converted area is 15931 km2.     

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。     

基于NDVI的新疆荒漠地区植被覆盖度遥感估算经验模型研究

新疆荒漠地区植被覆盖度遥感估算模型十分缺乏,给荒漠化监测等相关工作带来很大不便,开展植被覆盖度遥感估算经验模型研究,对于促进和完善相关地区的生态监测及研究工作具有积极的现实意义。通过对阜康市北部沙漠南缘和克拉玛依市中部平原荒漠进行无人机航拍,利用无人机遥感提取(光合)植被信息,并将无人机航拍影像的植被覆盖度统计单元与高分辨率卫星影像像元在空间上直接相对应,获取在高分辨率卫星影像像元尺度上的植被盖度,然后通过植被覆盖度和空间上与其相对应的源自高分辨率卫星影像的NDVI数据的拟合关系,建立基于源自高分二号影像的NDVI的阜康北部沙漠植被覆盖度遥感估算线性模型以及基于源自ZY1-02C影像的NDVI的克拉玛依平原荒漠植被覆盖度遥感估算二次多项式模型。研究中所采用的无人机遥感与卫星遥感相结合、植被覆盖度统计单元与卫星像元在空间上直接对应的方法,可避免以往相关工作中常以点位测量数据代表卫星像元数据所带来的不确定性。由于所用卫星影像的NDVI数据稳定性相对不足等原因,所建立的遥感估算模型的估算精度尚相对偏低,有待于今后进一步的工作加以改进。     

基于遥感的黄土高原林草植被变化对河川径流的影响分析

从黄土高原不同地貌区降雨产流机制入手,分析了林草植被影响流域水循环的可能环节;利用20世纪70年代以来不同时期的土地利用和植被盖度解译成果,以及同期实测的降雨和径流数据、供用水数据等,引入林草植被覆盖率、径流系数、产洪系数和基流系数等概念,从流域尺度上构建了林草植被覆盖率与河川径流的定量响应关系,结果发现,在半湿润或半干旱的黄土区,径流系数和产洪系数都将随林草植被的改善而减少,气候越干旱、径流或洪量减少越多;与同气候带的黄土区相比,盖沙黄土区林草植被改善所导致的减水量更大。不过,当林草植被覆盖率大于60%后,产洪系数变化减缓;最终河川径流将稳定在大于基流的某阈值附近。     

Spatial structure analysis of a reptile community with airborne LiDAR data

The analysis of the spatial structure of animal communities requires spatial data to determine the distribution of individuals and their limiting factors. New technologies like very precise GPS as well as satellite imagery and aerial photographs of very high spatial resolution are now available. Data from airborne LiDAR (Light Detection and Ranging) sensors can provide digital models of ground and vegetation surfaces with pixel sizes of less than 1 m. We present the first study in terrestrial herpetology using LiDAR data. We aim to identify the spatial patterns of a community of four species of lizards (Lacerta schreiberi, Timon lepidus, Podarcis bocagei, and P. hispanica), and to determine how the habitat is influencing the distribution of the species spatially. The study area is located in Northern Portugal. The position of each lizard was recorded during 16 surveys of 1 h with a very precise GPS (error < 1 m). LiDAR data provided digital models of surface, terrain, and normalised height. From these data, we derived slope, ruggedness, orientation, and hill-shading variables. We applied spatial statistics to determine the spatial structure of the community. We computed Maxent ecological niche models to determine the importance of environmental variables. The community and its species presented a clustered distribution. We identified 14 clusters, composed of 1–3 species. Species records showed two distribution patterns, with clusters associated with steep and flat areas. Cluster outliers had the same patterns. Juveniles and subadults were associated with areas of low quality, while sexes used space in similar ways. Maxent models identified suitable habitats across the study area for two species and in the flat areas for the other two species. LiDAR allowed us to understand the local distributions of a lizard community. Remotely sensed data and LiDAR are giving new insights into the study of species ecology. Images of higher spatial resolutions are necessary to map important factors such as refuges.     

Supporting multi-resolution out-of-core rendering of massive LiDAR point clouds through non-redundant data structures

In recent years, the evolution and improvement of LiDAR (Light Detection and Ranging) hardware has increased the quality and quantity of the gathered data, making the storage, processing and management thereof particularly challenging. In this work we present a novel, multi-resolution, out-of-core technique, used for web-based visualization and implemented through a non-redundant, data point organization method, which we call Hierarchically Layered Tiles (HLT), and a tree-like structure called Tile Grid Partitioning Tree (TGPT). The design of these elements is mainly focused on attaining very low levels of memory consumption, disk storage usage and network traffic on both, client and server-side, while delivering high-performance interactive visualization of massive LiDAR point clouds (up to 28 billion points) on multiplatform environments (mobile devices or desktop computers). HLT and TGPT were incorporated and tested in ViLMA (Visualization for LiDAR data using a Multi-resolution Approach), our own web-based visualization software specially designed to work with massive LiDAR point clouds.     

Effectiveness of farmland transfer in alleviating farmland abandonment in mountain regions

Farmland abandonment is a type of land use change in the mountain region, and this change is under rapid development. Whether farmland transfer can prevent this process and promote the effective allocation of land resources or not is a question worth studying and discussion. With the help of the previous research findings, the objective of this paper was to find out the role of farmland transfer on preventing farmland abandonment, by using the methods of multiple view with two factors, and single factor correlation analysis. The results showed that: (1) At village level, a significant negative correlation between farmland transfer and farmland abandonment existed in the study site, with R² = 0.7584. This correlation of farmland with high grade farming conditions presented more outstandingly. The fitted curve for the farmland at Level I had the largest R² at 0.288, while that for the farmland at Level IV had the smallest R² at 0.103. Which indicated that farmland transfer could prevent the abandonment of farmland with high grade farming conditions? (2) At plot level, the abandonment rate of farmland with high grade farming conditions was significantly lower than that of farmland with poor grade farming conditions. It was the lowest at 10.49% for the farmland with Level I farming conditions, whereas the farmland with Level I farming conditions was 26.21%. Abandoned farmland was mainly contributed by farmland with Level IV farming conditions in the study site. (3) At village level, the role of farming conditions on farmland abandonment was insignificant. The univariate correlation analysis revealed that the abandonment ratio was negatively correlated with the proportions of farmland at Levels I and II and their accumulated proportion; however, their R² were small at 0.194, 0.258, and 0.275, respectively. The abandonment of farmland with high farming conditions still existed. The abandonment ratios of farmland at Levels I and II were high at 9.96% and 10.60%, respectively. This presented that farmland transfer on behalf of the land rental market was still not developed. (4) However, the village possessed the high rate of farmland transfer, and its rate of farmland abandonment with high grade farming conditions was all lower. When the transfer ratios of farmland were over 20%, the abandonment ratios of farmland at Levels I and II were 6.47% and 6.92%, respectively. Farmland abandonment was still controlled by the improvement of land rental market. And the functions of land rental market optimizing the utilization of farmland resources have been presented to a certain degree. (5) To further improve the marketing degree of land rental, the probability of farmland abandonment could be reduced. Especially, their function to farmland with high grade farming conditions was very obvious, and could avoid the waste of farmland resources with high grade farming conditions.