基于决策树分类的大屯矿区地物信息提取及矿区污染分析

本文在对国内外遥感图像分类方法充分研究分析的基础上,选择决策树分类法对大屯矿区的Landsat 8遥感图像进行分类研究。选取样本提取并分析研究区典型地类光谱特征曲线,依据光谱曲线特征和归一化植被指数建立了土地利用分类决策树模型,通过反复试验和修正,筛选出适宜大屯矿区地物分类的决策树最优阈值,对研究区进行分类和精度评价,最后通过分类结果对研究区的水体污染状况进行简要分析。     

风暴分类识别技术在人工防雹中的应用

利用新一代多普勒天气雷达资料,在风暴跟踪识别算法的基础上,发展了风暴分类技术,以提高人工防雹作业指挥的效率。首先以SCIT算法为基础,结合风暴的结构特征,综合利用雷达、探空资料,自动提取风暴结构特征指数;其次采用基于决策树模型的风暴自动分类技术,将风暴按强度分为雷雨云、单体风暴、多单体风暴和强风暴;最后根据风暴强度、高度和位置等属性,对有可能产生冰雹的单体,结合GIS,自动对下游方向或附近作业点进行预警或输出作业参数。通过对2006—2014年期间重庆、辽宁大连和河南三门峡三地发生的较为典型的31次冰雹天气过程、182站次冰雹样本的检验来看:该方法通过对风暴按强度、垂直结构等综合属性进行分类,能有效提高冰雹识别的命中率、降低空报率,其中强风暴的命中率能达到100%,空报率仅为11.4%。能有效提高人工防雹作业的自动化程度,对防雹作业的科学决策有着重要参考作用。     

顾及位置关系的网络POI地址信息标准化处理方法

针对互联网POI(兴趣点)地址信息中广泛存在的地址要素不完整、文字表达不一致等不规范现象,提出一种顾及位置关系的网络POI地址信息标准化处理方法,首先对POI信息进行切分提取并逐层匹配地址树模型;然后基于4种位置关系从标准POI库中选出相应集合,作为丰富和修正非标准POI地址要素的候选;最后通过最小粒度地址要素的回溯,实现POI地址信息的快速标准化处理。试验表明该方法可以获得较高的准确率,尤其适用于在互联网数据环境中的POI地址信息标准化。     

回归-Elman网络在矿坝变形预测中的应用

金矿开展矿坝的变形监测工作,引入多层递阶回归分析模型,有较高的预测精度,但方法较繁琐,计算较复杂。由于变形数据可分离成趋势项与随机项,趋势项可用多元线性回归良好地拟合;随机项的预测,文中采用Elman网络建模计算,最后利用矿坝的实测高程位移数据进行验证,并与多层递阶回归进行比较。结果表明:回归-Elman网络模型比多层递阶回归的预测精度更高,效果更好,且方法简洁实用性强。     

改进后灰色神经网络及短期基坑沉降预测研究

在短期基坑沉降监测中,由于数据量少且呈非线性变化,沉降模型很难准确建立。灰色GM(1,1)对数据少、趋势性强、波动小的数据有较高的预测精度,但不能模拟复杂的非线性函数;BP神经网络可以对非线性数据进行学习训练,具有自学习、自适应能力;通过将GM(1,1)与BP神经网络组合,并优化网络部分的学习率、权值和阈值等,建立一种改进的灰色神经网络模型,该模型具有对非线性数据自学习、自适应能力和预测精度更高等优点。通过某基坑沉降监测分析,验证改进的灰色神经网络模型预测精度更高,适合短期建模,具有很好的实用性。     

一种裸露土壤湿度反演方法

针对目前土壤湿度反演方法研究较少且缺少实时性的现状,该文提出一种土壤湿度反演方法——最小二乘支持向量机技术。以积分方程模型为正向算法,数值模拟不同雷达参数(频率、入射角及极化)下后向散射系数随土壤含水量和地表粗糙度的变化情况。经过数据敏感性分析,选取C-波段和X-波段、小入射角下的同极化后向散射系数作为支持向量回归的训练样本信息;经过适当的训练,利用支持向量回归技术对土壤含水量进行了反演研究;并考虑通过多频率、多极化、多入射角数据的组合,消除地表粗糙度的影响,提高反演精度。模拟结果表明,该方法反演土壤湿度具有较高的精度和较好的实时性;同时,与人工神经网络方法的结果比较,证明了该方法的有效性,为土壤湿度的反演研究提供了一种方法。     

高维遥感图像的快速分类算法

为了实现对高维遥感图像的快速准确分类,提出了一种基于k均值二叉树支持向量机(SVM)的分类方法。该方法通过对选取的训练样本进行k均值聚类,生成支持向量机分类二叉树,作为确定最佳分类顺序的依据,以降低分类过程中的误差累积并提高整体分类精度,而且可缓解由样本数量不均衡导致的分类误差。该方法可在不进行降维处理的情况下,对高维遥感图像进行快速准确分类。测试结果表明,其分类速度和分类精度都优于传统的支持向量机分类结果。     

Validation of MODIS Vegetation Continuous Fields for monitoring deforestation and forest degradation: two cases in Mexico

This study assesses whether MODIS Vegetation Continuous Fields percent tree cover (PTC) data can detect deforestation and forest degradation. To assess the usefulness of PTC for detecting deforestation, we used a data set consisting of eight forest and seven non-forest categories. To evaluate forest degradation, we used data from two temperate forest types in three conservation states: primary (dense), secondary (moderately degraded) and open (heavily degraded) forest. Our results show that PTC can differentiate temperate forest from non-forest categories (p = 0.05) and thus suggests PTC can adequately detect deforestation in temperate forests. In contrast, single-date PTC data does not appear to be adequate to detect forest degradation in temperate forests. As for tropical forest, PTC can partially discriminate between forest and non-forest categories.     

Integrating cellular automata,artificial neural network,and fuzzy set theory to simulate threatened orchards: application to Maragheh,Iran

Urbanization processes challenge the growth of orchards in many cities in Iran. In Maragheh, orchards are crucial ecological, economical, and tourist sources. To explore orchards threatened by urban expansion, this study first aims to develop a new model by coupling cellular automata (CA) and artificial neural network with fuzzy set theory (CA–ANN–Fuzzy). While fuzzy set theory captures the uncertainty associated with transition rules, the ANN considers spatial and temporal nonlinearities of the driving forces underlying the urban growth processes. Second, the CA–ANN–Fuzzy model is compared with two existing approaches, namely a basic CA and a CA coupled with an ANN (CA–ANN). Third, we quantify the amount of orchard loss during the last three decades as well as for the upcoming years up to 2025. Results show that CA–ANN–Fuzzy with 83% kappa coefficient performs significantly better than conventional CA (with 51% kappa coefficient) and CA–ANN (with 79% kappa coefficient) models in simulating orchard loss. The historical data shows a considerable loss of 26% during the last three decades, while the CA–ANN–Fuzzy simulation reveals a considerable future loss of 7% of Maragheh’s orchards in 2025 due to urbanization. These areas require special attention and must be protected by the local government and decision-makers.     

A Survey of the Spatio-Temporal Data Model

Traditional GIS(Geographic Information System)mostly can only describe the transient state data and does not have the ability to deal with the temporal dynamic data. When the data changes, new data will take the place of the old data, namely the transient state change into another, and the old data will disappear. Therefore, it is unable to make an analysis of the updating changes of the data and predict the development trend of the future. In this case, TGIS (Temporal Geographic Information System) emerges and expands the time dimension on the basis of traditional GIS. Spatio-temporal data model is the key to TGIS. Spatio-temporal data modeling is not only related to dynamic expression of spatio-temporal objects, but also gives an important support for spatio-temporal analysis and reasoning. This paper summarized the theories and applications status at home and abroad of spatio-temporal data model in detail, illustrated family tree of spatio-temporal data model for the nearly fifty years, discussed improvement and application status of Base State with Amendments Model, Event-based Spatio-temporal Data Model, Object-oriented Data Model and other spatio-temporal data models, and raised the existing problems of spatio-temporal data model. The current existing problems mainly includes: ①There are a lot of spatio-temporal data models put forward, but some of them only focus on semantic design and neglect the verified; ②Most of existing spatio-temporal data model are for vector data, only the Event-based Spatio-temporal Data Model is raster data structure; ③At present, the expression the time-space information of geographic entity is relative separated with spatio-temporal data model; ④Spatio-temporal data model is mainly used in cadastral management, land use and forestry data updating, less application in other fields. In the end, the future development direction was put forward of spatio-temporal data model. In the time of big data and “Internet plus”, it is necessary to explore the big data spatio-temporal data model that supports multiple data formats.