支持向量机的车载雷达点云目标识别

为了进一步研究移动测量系统的数据处理问题,该文根据点云的基本特征,归纳了由7个特征构成的点云原始特征向量,在此基础上,结合语义环境构建了由17个特征构成的点云扩展特征向量,并采用支持向量机模型对车载LiDAR点云进行行道树点云识别的一系列实验。实验中采用粒子群优化算法和遗传算法对支持向量机进行参数寻优;采用不同特征向量和不同数目样本对点云进行学习和目标识别;分析了特征向量的学习曲线和识别精度。实验结果表明,支持向量机模型能够在行道树点云识别中取得较高的精度。     

融合机载雷达和航空正射影像的道路提取

针对单一遥感数据源提取道路信息中存在的问题,该文提出一种融合多源数据的方法:采用面向对象思想,融合机载激光雷达和航空正射影像,运用对象平均强度指数进行道路精确提取。首先利用点云强度信息进行道路粗提取,再结合影像光谱信息对粗提取点云进行优化,得到"一次点云";然后通过分形网络演化算法对影像进行多尺度分割形成对象,将"一次点云"导入分割对象,利用对象平均强度辅之以对象面积完成道路精确提取。实验结果表明,该方法提取流程简单快速、无须精确配准、道路提取精度较高。     

基于点云空洞修复和TPS模型的数学形态学机载LIDAR滤波

在分析现有滤波算法的基础上,结合数据驱动和模型驱动式算子各自的优点,提出基于点云空洞修复和TPS变形模型的数学形态学机载LIDAR点云滤波,该方法首先提取和修复由水域造成的大面积点云空洞,采用多尺度形态学开算子作用于修复的数据,得到近似裸露地表面;然后利用2D空间的TPS变形模型,以近似地表面为基础,插值原始点云,根据插值与原始点云高程的差值大小去识别地面点和非地面点。通过定量分析,验证该方法不仅有较高的滤波精度,而且也能较好的保留裸露地表的细节特征,同时该方法有助于辅助人工处理,提高数据处理的质量。     

基于可移动角点的航空和地面LiDAR数据配准

针对航空和地面LiDAR数据配准中点云数据的共轭特征较少且精度差异较大的问题,提出了一种基于可移动角点的航空和地面LiDAR数据配准方法:从航空和地面LiDAR数据中分别提取相应的建筑物角点,采用6参数模型对角点进行初始配准;以地面角点为参照,利用迭代移动方法对误差较大的航空角点进行修正;最后根据移动后的航空和地面角点计算获得点云配准关系。实验结果表明,该文方法可取得较好的点云配准效果,角点修正后能有效提升点云配准精度,适合于含有角点特征的航空和地面LiDAR数据配准。     

Opportunistic natural experiments using digital telemetry: a transit disruption case study

ABSTRACT

In the past decade society has entered a technological period characterized by handheld computing that supports input and processing from numerous sensors. Today’s mobile phones offer the ability to integrate input from sensors monitoring various external and internal sources (e.g., accelerometer, magnetometer, microphone, GPS, wireless Internet, and Bluetooth). Furthermore, these raw inputs can be integrated and processed in ways that can offer novel representations of human behaviour. As a result, new opportunities to examine and better understand human spatial behaviour are available; one such application is the constant monitoring of a group of people over an extended period of time. Such a research setting lends itself to natural experiments that emerge as a result of regular and on-going observations. We report here on the observation of a natural experiment that took place in the context of a month-long monitoring study of 28 participants using mobile phone-based ubiquitous sensor monitoring. The implications for public health and transportation planning are discussed.     

Evaluation of effectiveness of three fuzzy systems and three texture extraction methods for building damage detection from post-event LiDAR data

Building damage maps after disasters can help us to better manage the rescue operations. Researchers have used Light Detection and Ranging (LiDAR) data for extracting the building damage maps. For producing building damage maps from LiDAR data in a rapid manner, it is necessary to understand the effectiveness of features and classifiers. However, there is no comprehensive study on the performance of features and classifiers in identifying damaged areas. In this study, the effectiveness of three texture extraction methods and three fuzzy systems for producing the building damage maps was investigated. In the proposed method, at first, a pre-processing stage was utilized to apply essential processes on post-event LiDAR data. Second, textural features were extracted from the pre-processed LiDAR data. Third, fuzzy inference systems were generated to make a relation between the extracted textural features of buildings and their damage extents. The proposed method was tested across three areas over the 2010 Haiti earthquake. Three building damage maps with overall accuracies of 75.0%, 78.1% and 61.4% were achieved. Based on outcomes, the fuzzy inference systems were stronger than random forest, bagging, boosting and support vector machine classifiers for detecting damaged buildings.     

Use of LiDAR-derived DEM and a stream length-gradient index approach to investigation of landslides in Zagros Mountains,Iran

This paper presents an approach to stream length-gradient index analysis to identify tectonic signatures. The graded profile of the Dez River in Zagros Mountains, Iran, indicates that the area has been tectonically disturbed, and it triggers landslide hazards. The high-gradient index shows that a steeper gradient could be potentially a signature for landslides identification. The digital surface models acquired by airborne LiDAR were used in this study to generate the HRDEM. Our result shows a great potential for improving landslide investigations by implementing stream length-gradient index derived from the HRDEM in conjunction with the landslide inventories data-set in the GIS environment. We also identified a correlation between the stream length-gradient index and the graded topographic profile with slopes and landslides. This empirical approach was verified by geodata analytics and landslide inventories data-set in conjunction with field observations. This study has identified the locations of high-gradient indices with susceptible to landslides.     

Numerical prediction of MODUs drift during Hurricane Katrina

Severe hurricanes, such as Katrina, broke the mooring lines of a number of mobile offshore drilling units (MODU) deployed in the Gulf of Mexico and some of those MODUs went adrift. A drifting MODU may damage other critical elements of the offshore oil and gas infrastructure by colliding with floating or fixed production systems and transportation hubs, or by rupturing pipelines owing to their dragging anchors over the seabed. To avoid or mitigate the damage caused by a drifting MODU, it is desirable to understand the mechanics of the drift of a MODU under the impact of severe wind, wave and current and have the capability of predicting the trajectory of the drift. To explore the feasibility and accuracy of predicting the trajectory of a drifting MODU based on hindcast met-ocean conditions and limited knowledge of the condition of the drifting MODU, this study employed a simplified equation describing only the horizontal (surge, sway and yaw) motions of a MODU under the impact of steady wind, current and wave forces. The simplified hydrodynamic model neglects the first- and second-order oscillatory wave forces, unsteady wind forces (owing to wind gustiness), wave drift damping, and the effects of the body oscillation on the steady wind and current forces. It was assumed that the net effects of the oscillatory forces on the steady motion are insignificant. To verify the accuracy and feasibility of our simplified approach, the predicted drifting trajectories of two MODUs were compared with the corresponding measurements recorded by the global positioning system (GPS).     

基于LiDAR数据的DEM和矢量自动提取探讨

近年来LiDAR技术发展迅速,应用潜力巨大.基于TerraSolid软件环境,简要介绍了LiDAR数据的分类,在此基础上,重点讨论了LiDAR数据的DEM和矢量自动提取的相关问题.由LiDAR数据生成的DEM精度较高.但矢量自动提取还未到实用阶段.     

Curvature derived from LiDAR digital elevation models as simple indicators of debris-flow susceptibility

To mitigate the damage caused by debris flows resulting from heavy precipitation and to aid in evacuation plan preparation, areas at risk should be mapped on a scale appropriate for affected individuals and communities. We tested the effectiveness of simply identifying debris-flow hazards through automated derivation of surface curvatures using LiDAR digital elevation models. We achieved useful correspondence between plan curvatures and areas of existing debris-flow damage in two localities in Japan using the analysis of digital elevation models(DEMs). We found that plan curvatures derived from 10 m DEMs may be useful to indicate areas that are susceptible to debris flow in mountainous areas. In residential areas located on gentle sloping debris flow fans, the greatest damage to houses was found to be located in the elongated depressions that are connected to mountain stream valleys. Plan curvaturederived from 5 m DEM was the most sensitive indicators for susceptibility to debris flows.