Automated Kerogen Classification in Microscope Images of Dispersed Kerogen Preparation

We develop the classification part of a system that analyses transmitted light microscope images of dispersed kerogen preparation. The system automatically extracts kerogen pieces from the image and labels each piece as either inertinite or vitrinite. The image pre-processing analysis consists of background removal, identification of kerogen material, object segmentation, object extraction (individual images of pieces of kerogen) and feature calculation for each object. An expert palynologist was asked to label the objects into categories inertinite and vitrinite, which provided the ground truth for the classification experiment. Ten state-of-the-art classifiers and classifier ensembles were compared: Naïve Bayes, decision tree, nearest neighbour, the logistic classifier, multilayered perceptron (MLP), support vector machines (SVM), AdaBoost, Bagging, LogitBoost and Random Forest. The logistic classifier was singled out as the most accurate classifier, with an accuracy greater than 90. Using a 10 times 10-fold cross-validation provided within the Weka software, we found that the logistic classifier was significantly better than five classifiers (p<0.05) and indistinguishable from the other four classifiers. The initial set of 32 features was subsequently reduced to 6 features without compromising the classification accuracy. A further evaluation of the system alerted us to the possible sensitivity of the classification to the ground truth that might vary from one human expert to another. The analysis also revealed that the logistic classifier made most of the correct classifications with a high certainty.     

基于知识的点状符号自动识别

文中根据地形图黑版要素中点状符号的特点,总结了6种易于表达的知识特征,并据此提出了一种基于知识的点状符号识别方法。     

断阶型潜山带地质模型的建立及储层预测--以CG20潜山为例

以 CG2 0潜山为例 ,从建立地质模型入手 ,包括地层模型、构造模型、储集模型、储盖组合模型、速度模型等 ,认识到各套地层分布和储层物性的差异均与地震响应密切相关 ,因此可以利用地震波的信息 ,预测潜山储层的发育及分布情况。在对 CG2 0潜山进行精细全三维构造解释的基础上 ,探讨性地应用了测井约束反演、吸收系数、相干分析及三维模式识别等技术 ,对潜山储层进行了预测 ,从而提高了潜山勘探的效益 ,并为类似断阶型潜山带的勘探提供了成功的经验 ,具有一定的指导意义。     

一种基于深度学习的城市道路积水内涝图像识别方法

为了提高超大城市积水内涝的监测预警能力,针对现有积水内涝检测方法实用性、实时性存在的不足,依托广州市9万余个高密度公共监控视频作为数据基础,基于深度学习的目标检测RTMDet (Real-Time Models for Object Detection) 模型建立了一种道路积水内涝图像识别方法,采用多线程技术实现大规模视频点图像的高效采集,同时基于实例分割RTMDet模型实现积水内涝目标的快速检测和识别。使用2018年8月15日—2020年5月23日通过视频接口和互联网收集的共8 463张图像样本集构建识别模型,用2020年6月25日—2022年9月10日采集的6 106张图像对模型进行检验评估。结果表明：强降水过程中,模型能准确识别出具有明显积水内涝特征的图像及积水内涝特征所处图像的位置区域,数据清洗后模型总体识别准确率为86.60%;光线干扰和降水造成摄像头画面模糊是导致模型虚警的两大主要因素。研究成果已经在广州气象影响预报业务检验和广州市应急指挥决策辅助系统应用,为城市积水内涝自动监测和预警提供有效支撑,对水务、交通等气象业务场景有一定的参考价值。     

东亚夏季风的年际变化及其对环流和降水的影响

Using NCEP/NCAR reanalysis geopotential height (GHT) and wind at 850 hPa, GHT at 500 hPa, precipitation rate, sea level pressure (SLP) and precipitation observations from more than 600 stations nationwide in June–August from 1951 to 2006, and focusing on the East Asia-West Pacific region (10°–80°N, 70°–180°E), interannual variation of East Asian summer monsoon (EASM) and its correlations with general circulation and precipitation patterns are studied by using statistical diagnostic methods such as 9-point high pass filtering, empirical orthogonal function (EOF) analysis, composite analysis and other statistical diagnosis, etc. It is concluded as follows: (1) EOF analysis of SLP in the East Asia-West Pacific region shows the existence of the zonal dipole oscillation mode (APD) between the Mongolia depression and the West Pacific high, and APD index can be used as an intensity index of EASM. (2) EOF analysis of GHT anomalies at 500 hPa in the East Asia-West Pacific region shows that the first EOF mode is characterized with an obvious meridional East Asian pattern (EAP), and EAP index can also be used as an EASM intensity index. (3) The composite analysis of high/low APD index years reveals the close correlation of APD index with EAP at 500 hPa (or 850 hPa). The study shows an obvious opposite correlation exists between APD index and EAP index with a correlation coefficient of −0.23, which passes the confidence test at 0.10 level. (4) Both APD and EAP indexes are closely correlated with precipitation during flood-prone season in China and precipitation rate over the East Asia-West Pacific region. The significant correlation area at 5% confidence level is mainly located from the southern area of the Yangtze River valley to the ocean around southern Japan, and the former is a positive correlation and the latter is a negative one. Foundation: Cooperative Project funded by the Ministry of Science and Technology of the People’s Republic of China, No.2007DFB20210; National Natural Science Foundation of China, No.90502003; JICA China-Japan Technical Cooperative Project “China-Japanese Cooperative Research Center on Meteorological Disasters”. Author: Yu Shuqiu, Associate Professor, specialized in climate and climate change.     

顾及形态特征和语义信息的双线道路识别方法

双线道路识别与提取是城市路网综合的关键。提出了一种城市双线道路提取的方法。在构建好的道路网眼的基础上,综合考虑双线道路网眼形态特征以及构成网眼路段之间的语义相似度,设计了识别双线道路网眼的综合指标。通过判断网眼综合指标确定最终的双线道路网眼,最后根据识别出的双线道路网眼提取出双线道路。实验表明,该方法能有效提取出城市道路数据中的双线道路。     

渤中A凝析气田潜山储层特征及其地球物理判识方法

渤中A凝析气田为典型的大型太古代变质岩潜山气田,岩性主要为片麻岩、变质二长花岗岩、变质花岗闪长岩和碎裂二长花岗岩,储集空间以裂缝-孔隙和孔隙-裂缝的组合类型为主.纵向上潜山划分为半风化带和潜山内幕,半风化带普遍发育网状缝和粒内溶蚀孔,形成储集性能好的块状储集体,潜山内幕非均质性强,储集性能差于半风化带;裂缝是有效的储集空间又是良好的渗流通道,大断层附近裂缝相对富集.孔隙型储层的储集空间多为粒间孔和溶蚀孔,主要发育在潜山顶部或潜山内幕断层附近;半风化带储层物性及其溶蚀孔隙展布与断裂密切相关,古地貌高点和断裂系统发育耦合区是有利的储层发育区带;裂缝的密度和开度对气井无阻流量和产量有重要影响.     

基于ALBERT模型的园林植物知识实体与关系抽取方法

园林植物知识图谱可为顾及区域适应性、观赏性和生态性等因子的绿化树种的选型提供知识支持.植物描述文本的实体识别及关系抽取是知识图谱构建的关键环节.针对植物领域未有公开的标注数据集,本文阐述了园林植物数据集的构建流程,定义了园林植物的概念体系结构,完成了园林植物语料库的构建.针对现有Word2vec、ELMo和BERT等语...     

利用Faster R-CNN进行立交桥自动识别与定位

立交桥结构的自动识别对道路网多尺度建模、空间分析和车辆导航具有重要意义.传统基于矢量数据的立交桥识别方法过分依赖人工设计的特征,对复杂场景的适应性较差.本文提出了一种基于目标检测Faster R-CNN神经网络模型的立交桥识别方法,该方法利用卷积神经网络学习立交桥样本的深层次结构特征,进而实现立交桥的自动识别与准确定位...     

Spatial pattern recognition by decomposition

The decomposition problem—the assignment of sample observations to component populations—is studied in a spatial context. The observations are spatially located and the assignment to component populations takes into account the value of each observation as well as the values of neighboring observations. Both parameter estimation and assignment rules use a new method that integrates a standard multivariate decomposition algorithm with nonlinear regression. The method is illustrated and tested with artificial data. The distribution of the trace component Cr₂O₃ in recent Lake Michigan sediments is, then, analyzed by the method. It yields a pattern of component populations that is correlated with the Lake's bottom structure and depositional environments. Study carried out as visiting scholar at the Department of Geological Sciences, Northwestern University, Evanston, Illinois U.S.A.