Geospatial data mining for digital raster mapping

We performed an in-depth literature survey to identify the most popular data mining approaches that have been applied for raster mapping of ecological parameters through the use of Geographic Information Systems (GIS) and remotely sensed data. Popular data mining approaches included decision trees or “data mining” trees which consist of regression and classification trees, random forests, neural networks, and support vector machines. The advantages of each data mining approach as well as approaches to avoid overfitting are subsequently discussed. We also provide suggestions and examples for the mapping of problematic variables or classes, future or historical projections, and avoidance of model bias. Finally, we address the separate issues of parallel processing, error mapping, and incorporation of “no data” values into modeling processes. Given the improved availability of digital spatial products and remote sensing products, data mining approaches combined with parallel processing potentials should greatly improve the quality and extent of ecological datasets.     

Building a model-based personalised recommendation approach for tourist attractions from geotagged social media data

When travelling, people are accustomed to taking and uploading photos on social media websites, which has led to the accumulation of huge numbers of geotagged photos. Combined with multisource information (e.g. weather, transportation, or textual information), these geotagged photos could help us in constructing user preference profiles at a high level of detail. Therefore, using these geotagged photos, we built a personalised recommendation system to provide attraction recommendations that match a user's preferences. Specifically, we retrieved a geotagged photo collection from the public API for Flickr (Flickr.com) and fetched a large amount of other contextual information to rebuild a user's travel history. We then created a model-based recommendation method with a two-stage architecture that consists of candidate generation (the matching process) and candidate ranking. In the matching process, we used a support vector machine model that was modified for multiclass classification to generate the candidate list. In addition, we used a gradient boosting regression tree to score each candidate and rerank the list. Finally, we evaluated our recommendation results with respect to accuracy and ranking ability. Compared with widely used memory-based methods, our proposed method performs significantly better in the cold-start situation and when mining ‘long-tail’ data.     

基于SSW点云数据的矢量地图平面精度自动校验方法

常规的矢量地图精度校验采用抽样与实地测量,外业工作量大,自动化程度低。针对这一问题,本文提出基于SSW激光点云数据的矢量地图平面精度自动校验方法。首先,使用车载激光扫描器获得道路两侧高精度点云数据,并对点云数据进行滤波、坐标转换和精度检验;其次,基于多特征识别算法,使用SWDY软件提取点云特征点线;最后,利用最近邻法搜索待检矢量图中的同类地物特征点线,并计算匹配点线对的中误差。以兴化城区为试验区,采用该方法检测该地区1：1000比例尺的矢量地图平面精度,试验结果显示,成功匹配了点云数据205个地物特征中的201个,矢量地图的总体中误差为0.26 m,且能够发现待检测矢量地图中的采集丢漏与明显错误。本文方法可以减少现有检测方法的野外实测工作量,增加检测样本数量,降低检测过程中的人为干扰因素,有效提升检测的可靠性与检测效率。     

一种基于点云法向量的基准特征提取与倾斜分析方法

针对目前将三维激光扫描技术应用于变形监测领域存在基准特征难以提取、点云数据分析缺乏适用的方法等问题,本文提出了一种基于点云法向量的基准特征提取与形变分析方法。首先利用局部平面拟合方法获得点云的法向量,并沿点云法矢方向探测基准点;然后利用三次B样条曲线对探测的正确基准点进行拟合;最后根据拟合曲线计算基准高程和对径点倾斜角分析基准特征形变信息。对某化工厂的罐体点云数据进行基准特征提取结果表明,该方法可以快速、全面地获取监测对象的整体信息,且能够正确分析监测对象的基准形变。     

一种保持矢量数据几何形状的数字水印算法

针对目前矢量地理空间数据数字水印技术的需求,从矢量地理空间数据特点分析出发,运用矢量地理空间数据线状、面状要素具有的冗余数据的特性,首先在矢量地理数据中增加冗余数据点,然后通过控制冗余点在线状要素上的移动来嵌入水印信息,实现了高保真的矢量地理空间数据数字水印算法,最后对嵌入水印后数据的可用性以及算法的鲁棒性进行了试验分析。     

将微粒群和支持向量机用于耕地驱动因子选择的研究

结合微粒群算法(PSO)具有执行速度快、受问题维数变化影响小的优点及支持向量机算法(SVM)结构风险最小化原理,构建了基于离散二进制微粒群(BPSO)与支持向量机的耕地驱动力因子选择方法,使用特征子集中确定的特征来训练支持向量回归机,用适应度函数来评价回归机的性能,指导BPSO的搜索。实验表明,该方法能有效地提取出耕地驱动因子的特征子集,从而降低了指标的维数,保留了关键信息,以获得知识的最小表达。     

中国地区水平地壳形变模型的建立

我国大陆板块是欧亚大陆板块的重要组成部分,其板块运动剧烈,地质构造复杂,是全球研究板块运动的重要地区。本文通过对各个块体分别建模和解算,得出结论:中国地区除东北块体相对于欧亚板块不动外,其他块体均有不同程度的形变运动,东部形变较小,西部形变较大;华北块体总体上以大约6mm/a的速率向东运动;华南块体以约5~8mm/a的速率向东南方向运动;新疆块体以约10~15mm/a的速率向北东方向运动;西藏块体也向北东方向运动,但运动速率从南到北逐步减少。     

矢量数据的一种分层动态组织方法

提出矢量数据分层组织的一种动态结构,在该结构中,单点、单字符或符号作为基元,以属性串组织基元,并分层叠置。采用区别于链式结构的动态通用串结构及C语言予以实现。     

Tropical land use land cover mapping in Pará (Brazil) using discriminative Markov random fields and multi-temporal TerraSAR-X data

Remote sensing satellite data offer the unique possibility to map land use land cover transformations by providing spatially explicit information. However, detection of short-term processes and land use patterns of high spatial–temporal variability is a challenging task.We present a novel framework using multi-temporal TerraSAR-X data and machine learning techniques, namely discriminative Markov random fields with spatio-temporal priors, and import vector machines, in order to advance the mapping of land cover characterized by short-term changes. Our study region covers a current deforestation frontier in the Brazilian state Pará with land cover dominated by primary forests, different types of pasture land and secondary vegetation, and land use dominated by short-term processes such as slash-and-burn activities. The data set comprises multi-temporal TerraSAR-X imagery acquired over the course of the 2014 dry season, as well as optical data (RapidEye, Landsat) for reference. Results show that land use land cover is reliably mapped, resulting in spatially adjusted overall accuracies of up to 79% in a five class setting, yet limitations for the differentiation of different pasture types remain.The proposed method is applicable on multi-temporal data sets, and constitutes a feasible approach to map land use land cover in regions that are affected by high-frequent temporal changes.     

高光谱遥感影像分类研究进展

随着模式识别、机器学习、遥感技术等相关学科领域的发展,高光谱遥感影像分类研究取得快速进展。本文系统总结和评述了当前高光谱遥感影像分类的相关研究进展,在总结分类策略的基础上,重点从以核方法为代表的新型分类器设计、特征挖掘、空间-光谱分类、基于主动学习和半监督学习的分类、基于稀疏表达的分类、多分类器集成六个方面对高光谱影像像素级分类最新研究进行了综述。针对今后的研究方向,指出高光谱遥感影像分类一方面要适应大数据、智能化高光谱对地观测的发展前沿,继续引入机器学习领域的新理论、新方法,综合利用多源遥感数据、多维特征空间互补的优势,提高分类精度、分类器泛化能力和自动化程度;另一方面要关注高光谱遥感应用的需求,突出高光谱遥感记录精细光谱特征的优势,针对应用需求发展有效的分类方法。