基于模式识别理论的卡钻类型判别分析

将模式识别理论中的支持向量机、Bayes判别分析以及多元回归分析应用于卡钻类型判别分析,建立了基于模式识别理论的卡钻判别模型。以近几年川东北卡钻数据为例进行了算例分析,结果表明：采用支持向量机、Bayes判别法及多元回归法对卡钻类型判别的结果与实际结果的误判率分别为1.92%,11.52%,61.54%。支持向量机（SVM）判别结果精度最高,但其判别方程式较为复杂,不能直观看出各分量对结果的影响程度;多元回归分析判别方程形式简单,可以直观表达各参量与卡钻之间的密切程度,但其判别精度较低;Bayes判别法计算精度介于二者之间,但其判别精度与判别式的个数密切相关。     

Automatic Identification of Textcoded Measurement Targets

A method is presented which makes a valuable contribution to automation in close range metrology. Successful identification of textcoded measurement targets has three essential characteristics: (1) large and easily expandable code depth; (2) robustness with respect to radiometric and geometric influences; and (3) rejection capability in case of low identification quality. The proposed method of text recognition of target labels is based on experience which has been gradually accumulated in the context of automated document analysis at the DaimlerChrysler Research Center.     

利用多目标地球化学数据识别第四纪沉积物类型——基于概率神经网络方法

针对利用多目标地球化学数据研究第四纪沉积物类型问题,提出了基于概率神经网络的分类识别模型,并给出地球化学特征指标选取、指标归一化、神经网络设置和训练的具体方法、步骤。在吉林省中西部松嫩平原应用表明,该方法识别出8类不同成因的第四纪沉积物,较好地解决了该区第四纪沉积物成因归属问题。概率神经网络模型对第四纪沉积物类型的识别能力远高于常规多元统计方法,且结构简单、训练快捷。     

对地质灾害隐患早期识别相关问题的认识与思考

泥石流物源识别与计算是科学评估泥石流规模、危害程度以及综合治理的基础,而传统的地面调查和光学遥感手段难以有效识别山区植被茂密覆盖下的泥石流物源。机载激光雷达（light detection and ranging,LiDAR）技术能有效去除植被获取真实的地表形态,为泥石流物源的识别提供了新的解决方案。以九寨沟震区的日则沟泥石流为例,基于高分辨率机载LiDAR数据结合震前卫星影像,开展泥石流物源识别研究,根据物源所处位置和在山体阴影图像上的色彩及纹理差异,将物源分为崩滑物源、坡面物源和沟道物源,并建立各类型物源的机载LiDAR识别标志与遥感解译方法。共解译出日则沟泥石流物源155处,总面积达1.06 km²,占流域总面积的31.56%,在此基础上分析了各类型物源的发育分布规律。为泥石流物源的精确计算提供理论参考和数据支撑,进一步服务于九寨沟震区泥石流的防治与风险评价。

     

传统村落景观基因信息链与自动识别模型构建——以陕西省为例

传统村落作为活的文化遗产,承载了大量历史记忆,是地域文化景观基因识别与模型构建研究的重要切入点。以陕西省71个国家级传统村落为例,基于景观基因理论建立陕西省传统村落景观基因识别体系,识别出陕西省传统村落景观基因特征;运用类型学原理和N级编码理论对景观基因进行编码,构建陕西省传统村落景观基因信息链并生成基因谱系;借鉴生物学中“胞-链-形”DNA碱基序列模型,提取出环境基因、建筑基因、农耕文化基因和宗族文化基因四个公共基因作为景观基因元（胞）,以村内道路系统作为基因链,构建传统村落景观基因DNA模型与自动识别模型,以此对传统村落的区位、类型、特征和文化基因进行自动识别。为传统村落景观基因信息有效传承与存储,以及乡村建设动态发展提供理论借鉴。     

Candidate precursors: pulse-like geoelectric signals possibly related to recent seismic activity in Japan

Telemetric network observations of pulse-like geoelectric charge signals using a vertical dipole buried under the ground were undertaken at various observation sites over a wide area in Japan from April 1996. From continuous records of the signals during the six months following that, we attempted to select anomalous signals, possibly related to seismic electric activity. Special attention was paid to the elimination of noise resulting from industrial and meteorological electric activity, comparison with other electromagnetic signals in the VLF band and the relation between the precursor period and the distance from the eventual main shock that occurred in Japan. Four candidate precursor electric signals, which were not contaminated by industrial and meteorological electric activity, were then selected, of which the second appeared before the Akita-ken Nairiku-nanbu earthquake swarm, for which the maximum M = 5.9 occurred on 1996 August 11, and the third and fourth before the Chiba-ken Toho-oki earthquake, M = 6.6, on 1996 September 11. A tentative qualitative model explaining why the candidate precursory signal is related to stress building up before an earthquake is presented in terms of the electrification of gases released from fracturing rocks immediately prior to the main shock.     

EXPLORING SPATIAL FAMILIARITY*

Spatial familiarity is a natural language concept that has no well-defined technical interpretation. We explore here some possible dimensions of the concept, provide subjective measurements of the dimensions, and evaluate their independence. The four dimensions examined are locational knowledge, visual recognition, name identification, and interaction frequency. We suggest that both cognitive and behavioral dimensions are embedded in the concept of spatial familiarity and must be part of the meaning of the concept and of any attempt to make it operational.     

基于结构模式的道路网节点匹配方法

道路网数据中微观结构的识别对于多尺度路网建模、步行导航等至关重要。复杂道路交叉口是重要的道路微观结构之一,针对目前道路复杂交叉口基于几何形状描述与图形匹配识别方法存在的不足,从复杂交叉口识别与化简的角度出发,提出了一种利用路段分类进行复杂道路交叉口识别与化简的方法。该方法首先通过点密度聚类的方法对道路交叉口进行定位,然后利用路段的规模、形状和属性等特征构建特征空间,将交叉口的识别作为一种区分主干路段与辅助路段的两类分类问题,利用支持向量机的方法对交叉口区域内的路段进行分类,从而完成交叉口的识别与化简。利用开放街道地图（Open Street Map）数据进行实验,结果表明,该方法能够有效地识别道路交叉口。     

利用多分辨率半边进行全球多分辨率DEM无缝表达

矢量化的复杂建筑平面图纸受各类图元的干扰,其空间布局识别难度非常高,然而在5G基站建设、智能家居以及AR/VR（augmented reality/virtual reality）中广为应用。提出了一种结合栅格图像和矢量表示的高性能融合对偶识别方法。首先,采用建筑平面图纸的栅格化表示,提取建筑图纸的主要发展方向,构建分割的若干个封闭区域空间。然后,采用建筑平面图纸的矢量化表示,利用空间识别的有向性和邻接特性,创新地提出使用半墙数据结构进行墙体几何位置和拓扑关系的计算,从而根据空间对偶原理高精度地整体重建墙体布局。最后,获取包含各类墙体布局的矢量建筑平面图数据集,并进行方法的实验验证。实验结果证明了所提出的先空间后墙体的融合对偶识别算法对于各类建筑模型类型均具有可用性和有效性。与传统先墙体后空间的识别方法相比,所提方法具有更高的鲁棒性,受特定建筑图纸类型的干扰更少。