Crowd sensing and spatiotemporal analysis in urban open space using multi-viewpoint geotagged videos

Increasing concern for urban public safety has motivated the deployment of a large number of surveillance cameras in open spaces such as city squares, stations, and shopping malls. The efficient detection of crowd dynamics in urban open spaces using multi-viewpoint surveillance videos continues to be a fundamental problem in the field of urban security. The use of existing methods for extracting features from video images has resulted in significant progress in single-camera image space. However, surveillance videos are geotagged videos with location information, and few studies have fully exploited the spatial semantics of these videos. In this study, multi-viewpoint videos in geographic space are used to fuse object trajectories for crowd sensing and spatiotemporal analysis. The YOLOv3-DeepSORT model is used to detect a pedestrian and extract the corresponding image coordinates, combine spatial semantics (such as the positions of the pedestrian in the field of view of the camera) to build a projection transformation matrix and map the object recorded by a single camera to geographic space. Trajectories from multi-viewpoint videos are fused based on the features of location, time, and directions to generate a complete pedestrian trajectory. Then, crowd spatial pattern analysis, density estimation, and motion trend analysis are performed. Experimental results demonstrate that the proposed method can be used to identify crowd dynamics and analyze the corresponding spatiotemporal pattern in an urban open space from a global perspective, providing a means of intelligent spatiotemporal analysis of geotagged videos.     

Spatio-temporal variation and statistical characteristic of extreme dry spell in Yellow River Basin, China

Drought is one of the most detrimental natural hazards in Yellow River Basin (YRB). In this research, spatio-temporal variation and statistical characteristic of drought in YRB is studied by using dry spell. Two extreme series, including annual maximum series (AMS) and partial duration series (PDS), are used and simulated with generalized extreme value (GEV), generalized Pareto (GP), and Pearson type III (PE3) distributions. The results show that the northern part is drier than the southern part of YRB. Besides, the maximum dry spell usually starts in October, November, and December. According to the trend analysis, mean maximum length of dry spell (MxDS) shows a negative trend in most stations. From the L-moments and Kolmogorov–Smirnov test method, it can be found that GEV model can better fit AMS while GP and PE3 can better fit PDS. Moreover, the quantiles from optimal model of AMS and PDS depict a similar distribution with values increases from south to north. The spatial distribution of scale and location parameters of GEV model for AMS shows a south-to-north gradient, while the distribution of shape parameter is a little irregularity. Furthermore, based on the linear correlation analysis, there is an evident linear relation between location and scale parameters with mean and standard variation of MxDS, respectively.     

资阳地区强降水气候分析及其短期预报模式

本文通过对资阳的大～暴雨、暴雨的分布、强度和出现时段等气候统计分析,根据各类天气系统在关键区发生、发展及其演变规律进行分型,并建立了资阳强降水的预报模式.     

A novel method for simultaneous analysis of particle size and mineralogy for Chang’E-5 lunar soil with minimum sample consumption

Science China Earth Sciences - The successful return of lunar soil samples from the northern Oceanus Procellarum by the Chang’E 5 (CE-5) mission has provided unprecedented ground-truth...     

2D地震数据规则化中随机稀疏采样方案

地震数据规则化是地震信号处理中一个重要步骤,近年来受到广泛关注的压缩感知技术已经被应用到地震数据规则化中。压缩感知技术突破了传统的Shannon-Nyqiust采样定理的限制,可以用采集的少量地震数据重构完整数据。基于压缩感知技术的地震数据规则化质量主要受三个因素影响,除了受地震信号在不同变换域的稀疏表达和11范数重构算法的影响外,极大地取决于地震道随机稀疏采样方式。尽管已有学者开展了2D地震数据离散均匀分布随机采样方式研究,但设计新的稀疏采样方案仍然很有必要。在本文中,我们提出满足Bernoulli分布规律的Bernoulli随机稀疏采样方式和它的抖动形式。对2D数值模拟数据进行四种随机稀疏采样方案和两种变换（Fourier变换和Curvelet变换）实验,对获取的不完整数据应用11范数谱投影梯度算法（SPGL1）进行重构。考虑到不同随机种子点产生不同约束矩阵R会有不同的规则化质量,对每种方案和每个稀疏采样因子进行10次规则化实验,并计算出相应信噪比（SNR）的平均值和标准偏差。实验结果表明,我们提出的新方案好于或等于已有的离散均匀分布采样方案。     

古气候事件对刚果扇浊积水道发育的控制作用

刚果扇是渐新世以来快速堆积的巨型深水扇体,浊积水道广泛发育,油气成藏条件优越。精细标定水道集中发育的年龄、期次及其与同期地质事件之间的关联,对认识水道发育的主控因素和分布规律具有重要意义。基于测井、地震和古生物资料的综合分析,将盆地北部渐新统—中新统划分为2个二级层序、8个三级层序;通过微体古生物带和海平面变化旋回精细确定了各层序界面的年龄;进而在层序格架内明确了刚果扇北部浊积水道的发育期次和演化规律,认为刚果扇在盆地北部可划分为3个演化期次,即渐新世晚期快速填充期、中新世早—中期稳定沉积期和中新世晚期大规模水道发育期。实验研究发现,盆地中新统孢粉化石由早到晚出现水藓孢属-栎粉属、光面球藻属-盘星藻属-单束松粉属、拟桫椤孢属-凸瘤水龙骨单缝孢属、粗肋孢属-水龙骨单缝孢属-拟桦粉属4种组合类型,表明盆地中新世气候变化频繁,且大规模水道发育期与降温事件及海平面快速下降期高度吻合。此外,3次大规模水道发育期与西非海岸3次构造抬升事件关系密切。由此认为,刚果扇浊积水道的发育直接受控于降温事件,而构造隆升是源头性的触发因素。     

区域人群状态的实时感知监控

 公众聚集场所人群高度聚集,流动性大,隐藏着巨大的安全隐患,时常发生群死群伤的拥挤踩踏等突发公共事件。针对现有以视频监控的人群分析,无法从空间视角掌握区域人群状态的时空格局,本文提出了面向人群分析的视频GIS框架,将视频数据映射至地理空间,在GIS环境下提取人群密度、人群运动矢量场等人群特征。通过分析人群运动矢量场可得到人群运动模式及各方向人群主体运动速率。最后,将视频监控系统与GIS进行有机集成,设计并实现了以视频与GIS协同的区域人群状态实时感知监控系统。实验结果表明,本系统可为大型集会活动的突发事件预防、人群疏导等提供决策依据。     

碳酸盐岩油气藏的断裂改造作用——以塔里木盆地为例

多旋回盆地油气藏形成后多经历复杂的构造改造作用,并对油气的保存与分布具有重要的控制作用。结合塔里木盆地油气藏实例分析,将碳酸盐岩油气藏断裂改造作用划分为断裂切割改造、断裂抬升改造、断裂褶皱改造、断裂再活动改造与断裂成岩封闭改造等5种类型。断裂改造方式主要有圈闭变化、盖层减薄、水力作用、压力变化与成岩变化等,并造成油气逸散与补充、水洗与降解及流体性质的变化等,从而导致碳酸盐岩油气藏与油气分布的复杂性。断裂改造作用研究对多旋回油气盆地勘探开发具有重要借鉴意义。     

机载激光雷达技术在地质灾害调查中的应用——以四川九寨沟7.0级地震为例

九寨沟7.0级地震诱发的地质灾害具有高位滑坡远程灾害的特点,传统地质灾害排查手段无法有效解决隐患的早期识别问题。本文采用机载激光雷达测量技术,快速获取九寨沟地震核心景区的激光点云数据。通过构建高精度数字高程模型(DEM)、数字正射影像图(DOM),建立三维地质灾害解译场景,利用数字地形分析、地形形态学分析和计算机图像识别等技术,综合开展九寨沟高位远程区域内隐蔽性强、随机性大的地质灾害隐患早期识别与分析。应用实践表明,机载激光雷达测量技术可以提高九寨沟地震灾区地质灾害隐患的早期识别能力,对进一步提高综合防灾减灾能力提供了一些可借鉴的思路。