共查询到19条相似文献，搜索用时 390 毫秒
1.
Two Dimensional Principal Analysis是新近提出的一种图像分析方法,并已在特征提取与人脸和物体识别中得到较好应用.由于2DPCA本身就具有数据压缩功能,在用于去除图像的空间相关性中,可实现对数据的压缩,尤其是对高光谱或多光谱遥感图像的压缩.实验结果表明,这类方法不仅有效,而且压缩性能良好. 相似文献
2.
从数字高程数据(DEM)中自动提取自然水系及汇水盆地,并在GIS平台中表示出来,是开展基于GIS的区域水资源勘查与评价研究工作的基础,具有重要的实际意义。这里提出了一种根据自然水系自动提取结果,圈定各级水系汇水盆地,并将其表示成GIS中矢量图形数据的算法。该算法包括以下四个进程:①以水流方向数据和自然水系的分级标识数据为依据,将同一级别每一条水系汇水盆地内的栅格自动识别出来;②根据不同水系汇水盆地栅格的标识数据,自动搜索汇水盆地的边界栅格;③确定不同水系汇水盆地的分界线,并计算分界线上各个节点的X-Y坐标;④建立不同水系汇水盆地空间实体的拓扑关系,生成GIS可以识别的矢量图形数据。用我国东天山地区的DEM数据进行了算法的试验研究,实验结果表明,该算法是可行的。 相似文献
3.
InSAR（合成孔径干涉雷达测量）是一种新型对地观测技术,由于其高精度、大区域、可穿透云雾、全天候、全天时的工作特性,在数字高程模型（DEM）提取以及地表形变监测方面有着独特的优势.在InSAR图像处理的过程中,相位解缠是至关重要的一步,相位解缠效果的好坏直接影响到高程图的精度.本文利用ASAR数据,在ROI_PAC软件下进行处理,对DEM提取流程进行简要的分析.ROI_PAC软件处理InSAR数据所应用的相位解缠算法是最经典的路径跟踪法——枝切法.对于该算法本文进行了详细的分析探讨,并针对算法中残差点搜索的过程提出了一种新的改进算法思想. 相似文献
4.
5.
6.
高分辨率遥感数据处理方法实验研究 总被引：10，自引：0，他引：10
近几年高分辨率遥感数据处理与应用研究越来越受到国内外学者的重视,已成为目前遥感应用研究领域的热点和难点之一。首先分析了基于像素的中分辨率卫星数据的遥感数据处理方法,包括:基于统计理论的分类方法和基于光谱信息的非参数理论方法的分类方法。在此基础上,针对高分辨率遥感图像的特点,分析了面向对象的高分辨率遥感数据处理方法。其中,图像分割是最关键的一步,即如何有效地提取出对象特征非常重要。最后,给出了高分辨率遥感数据处理方法相关算法实验,主要进行了几种图像分割算法实验,同时结合植被变化检测,进行了变化检测算法实验。 相似文献
7.
遥感图像中地表水体同山体、建筑物等地物产生的阴影在光谱特征上存在较高的类间相似性,导致提取过程中容易出现混淆和错分的情况。针对此问题,提出一种基于面向对象和人工蜂群的地表水体提取方法。该方法首先对遥感图像进行分割以获取分割对象的光谱、比率、几何形状等统计特征,以弥补高分遥感图像波段数目少,信息量不足的缺陷;并借助人工蜂群算法在解决复杂问题最优化方面的优势,选取水体同阴影二值分类的几何平均正确率作为算法的适应度函数,最终获取地表水体的最优化提取规则。选取厦门市大嶝岛和湖南省资兴市部分区域,基于国产高分一号、二号遥感数据进行水体提取,并与传统SVM分类结果进行比较。实验结果表明本算法提取水体的总体精度和Kappa系数均优于传统SVM分类器,表明该方法可应用于高分遥感图像的地表水体提取。 相似文献
8.
9.
针对无人机遥感影像的特点,提出了一种基于特征点和重叠度的快速自动拼接算法.该算法实现了用于计算参考图像和待配准图像之间重叠度的相位相关法,并提出只在重叠区域中进行特征提取和特征匹配的方法.利用岷江流域无人机遥感图像进行了实验,结果表明,该算法使图像拼接有效地防止了图像非重叠区域中信息对算法的干扰,提高了拼接算法的精度和速度.并利用拼接的图像对茂县至都江堰的岷江主干河流进行了灾害信息提取和统计分析,为建立无人机高分辨率低空遥感数据的快速处理和应急灾害信息提取具有重要的指导意义. 相似文献
10.
基于成分关联区域相似度的面实体模糊匹配算法 总被引：1，自引：0，他引：1
空间目标匹配是空间数据库增量更新的第一步,也是关键一步.研究了基于空间目标匹配的变化信息的获取算法.通过研究空间数据中存在的不确定性问题,提出将模糊理论引入到空间目标匹配算法中.重点研究如何用模糊的方法解决空间目标匹配问题,并以面实体为例说明了具体匹配过程,提出了基于成分关联区域相似度的面实体模糊匹配算法.该算法利用成分关联区域的度量因子,确定模糊拓扑关系隶属度矩阵,进而量化隶属度矩阵,最终确定模糊拓扑关系分类.算法综合利用了图幅索引、成分关联因子等进行优化,简化计算复杂度,提高了算法效率. 相似文献
11.
以滇东北地区为例，通过对遥感地质异常的综合信息矿产预测方法进行了研究。在滇东北地区对ETM遥感图像进行岩性及构造的解译、遥感光谱异常提取等手段获得7项遥感地质异常信息因子，在GIS软件中，利用缓冲带（buffer zone）等方法将线性地质异常转化为面状异常，并通过空间分析确定各种遥感异常与已知矿产当量的关系，计算各主因子的二级因子得分，形成成矿预测栅格图像。通过栅格图像各通道间的复合运算分析进行遥感综合异常的定量圈定，最终在工作区圈定了多个不同级别的遥感综合异常。 相似文献
12.
Mingyue Lu Min Chen Xinhao Wang Manzhu Yu Yongyao Jiang Chaowei Yang 《Environmental Earth Sciences》2018,77(24):804
Weather radar data, which have obvious spatial characteristics, represent an important and essential data source for weather identification and prediction, and the multi-dimensional visualization and analysis of such data in a three-dimensional (3D) environment are important strategies for meteorological assessments of potentially disastrous weather. The previous studies have generally used regular 3D raster grids as a basic structure to represent radar data and reconstruct convective clouds. However, conducting weather radar data analyses based on regular 3D raster grids is time-consuming and inefficient, because such analyses involve considerable amounts of tedious data interpolation, and they cannot be used to address real-time situations or provide rapid-response solutions. Therefore, a new 3D modelling strategy that can be used to efficiently represent and analyse radar data is proposed in this article. According to the mode by which the radar data are obtained, the proposed 3D modelling strategy organizes the radar data using logical objects entitled radar-point, radar-line, radar-sector, and radar-cluster objects. In these logical objects, the radar point is the basic object that carries the real radar data unit detected from the radar scan, and the radar-line, radar-sector, and radar-cluster objects organize the radar-point collection in different spatial levels that are consistent with the intrinsic spatial structure of the radar scan. Radar points can be regarded as spatial points, and their spatial structure can support logical objects; thus, the radar points can be flexibly connected to construct continuous surface data with quads and volume data with hexahedron cells without additional tedious data interpolation. This model can be used to conduct corresponding operations, such as extracting an isosurface with the marching cube method and a radar profile with a designed sectioning algorithm to represent the outer and inner structure of a convective cloud. Finally, a case study is provided to verify that the proposed 3D modelling strategy has a better performance in radar data analysis and can intuitively and effectively represent the 3D structure of convective clouds. 相似文献
13.
基于影像的多源地学数据融合处理及应用 总被引：1，自引：0，他引：1
针对地学数据空间性、多源性和多解性的特点,提出了基于影像色觉原理的多源地学数据融合方法.融合的关键是地学信息的图像化,在云南镇康县芦子园地区多源地学数据融合处理中,根据不同时代地层对成矿的贡献,按重要性量化为256级的灰度值,实现地层数据的图像化;物化探异常通过插值生成栅格数据实现图像化;然后,利用图像化的地层、重力、化探、线环构造和遥感影像数据进行RGB合成,通过HIS变换进行多源地学数据融合.融合后的图像能充分显示研究区地层-构造-岩浆复合控矿特征,对进一步的找矿勘探有重要意义. 相似文献
14.
基于光谱相似尺度的支持向量机蚀变信息提取 总被引：4，自引：0，他引：4
文章提出一种基于光谱相似尺度(spectral similarity scale-SSS)的支持向量机(support vector machines-SVM)遥感数据矿化蚀变信息提取的新方法.该方法选择青海两兰地区作为遥感矿化蚀变信息典型研究区,利用该区域的Landsat7ETM遥感影像结合地面实况调查数据,从图像上选取少量具有代表性的样本点的光谱作为参考光谱,利用SSS方法提取训练样本,然后应用SVM算法进行遥感矿化蚀变信息提取.试验结果经野外检查和验证,效果良好. 相似文献
15.
卷积神经网络在图像识别领域处于领先地位,在目标检测方面的应用也越来越广泛,其特点是可以提取点与点之间的相关关系。二维地质图中,点与点之间往往存在特定的空间关系,将卷积神经网络技术应用于地质异常信息的提取有其重要性。在讨论卷积神经网络以及基于该技术目标检测算法(YOLO)的基础上,以湖南香花岭地区为例,提取与锡矿成矿相关的构造信息并进行分析,结果该方法能够覆盖原有矿点,有效地定义点与点之间的相关关系,描述点与点之间的空间相关性,可靠地提取与成矿有关的构造线密度信息,在花岗岩体与构造复杂地区圈出地质异常信息。 相似文献
16.
一种重磁特征线的自动提取方法 总被引：1，自引：0，他引：1
本文介绍一种最大梯度追踪和旋转变换方法，可用于重磁数字图像特征线自动提取。该方法分两步进行，首先用最大梯度追踪法产生二态像素图，然后用旋转变换方法将二态像素图的栅格特征线转换成矢量特征线。此方法生成的文件可以与ＧＩＳ接口。 相似文献
17.
An important issue in reservoir modeling is accurate generation of complex structures. The problem is difficult because the connectivity of the flow paths must be preserved. Multiple-point geostatistics is one of the most effective methods that can model the spatial patterns of geological structures, which is based on an informative geological training image that contains the variability, connectivity, and structural properties of a reservoir. Several pixel- and pattern-based methods have been developed in the past. In particular, pattern-based algorithms have become popular due to their ability for honoring the connectivity and geological features of a reservoir. But a shortcoming of such methods is that they require a massive data base, which make them highly memory- and CPU-intensive. In this paper, we propose a novel methodology for which there is no need to construct pattern data base and small data event. A new function for the similarity of the generated pattern and the training image, based on a cross-correlation (CC) function, is proposed that can be used with both categorical and continuous training images. We combine the CC function with an overlap strategy and a new approach, adaptive recursive template splitting along a raster path, in order to develop an algorithm, which we call cross-correlation simulation (CCSIM), for generation of the realizations of a reservoir with accurate conditioning and continuity. The performance of CCSIM is tested for a variety of training images. The results, when compared with those of the previous methods, indicate significant improvement in the CPU and memory requirements. 相似文献
18.
Vector and raster are two types of spatial data structures used in a geographic information system (GIS). With the development of GIS and remote sensing (RS) technologies, how to rapidly convert raster to vector data and establish topological relations among vectorized polygons is becoming a bottleneck in data integration between GIS and RS. Based on the previous work, an improved vectorization method is proposed to vectorize classified RS raster data quickly and automatically establish topological relations. In accordance with the connection information of arcs and nodes and both-sides polygonal attributes of arcs, the next arc can be searched directly by attribute matching when constructing polygons, thereby improving search efficiency. Moreover, our method addressed the problems of self-intersecting polygons, shared-boundary, and multi-nested islands and gave corresponding solutions, which can establish the topological relations of an entire image quickly. Two experiments, one for comparison between before and after vectorization of two different classified RS raster maps, and the other for comparison with several methods, are carried out to test the accuracy and efficiency of our method. Results show that the method solves the self-intersecting polygons, shared-boundary, and multi-nested islands problems. In addition, its vectorization speed is more than double that of commercial software ArcGIS, and the advantage of our method becomes more obvious as the number of polygons increases. Thus, our method can vectorize large and complex classified RS raster data with sufficient efficiency for practical use and establish topological relations among vectorized polygons. 相似文献
19.
对线性体图像进行分析,提出了一种基于扫描图像的线性体统计分析方法。通过对节点及其信息进行提取,快速抓住线性体的总体拓扑结构,并以节点信息为指引对线性体进行跟踪和识别,解决了复杂的线性体交叉问题。对识别出的线性体,通过直线拟合的方法,精确计算得到其长度、方位等信息。算法的主要特点是直接对扫描图进行运算,不需要矢量化,充分利用栅格图像的特点实现信息的快速获取,算法还具有精度高、抗噪性好和适应性强的特点。 相似文献