证据推理的遥感图像分类方法与应用

遥感图像分类一直是遥感研究领域的核心问题。然而,传统遥感分类方法在地物复杂地区不能取得满意效果。不但分类精度不够理想,分类灵活性也存在不足。本文尝试引入证据推理"软"分类方法,选择小兴安岭山区一景TM遥感图像,基于用户知识和经验,通过人机交互处理,以累计信任度(CBV)最大为划分像元归类的原则得到证据推理方法的分类图像。结果表明,整体分类精度从最大似然法的78.74%提高到82.28%,kappa系数从0.67提高到0.71。但该方法对于裸地分类精度不高,通过人为设定各类别CBV阈值的方法,获得了人为干预的证据推理方法分类图,其整体分类精度达到了87.80%,kappa系数也达到了0.81,所有类别的生产者和用户精度相比于最大似然法都有提高。研究表明,证据推理方法在遥感分类精度和分类灵活性方面都具有优越性。     

基于语义描述的制造业Web服务合成研究

在因特网环境下,如何对服务描述是基于web服务的应用集成的关键问题之一。Web服务合成技术为这一问题提供了很好的解决方案。以基于语义的web服务合成理论为基础,在分析制造业信息资源特点的基础之上,以库存管理领域为应用实例,选取面向服务的网络本体语言提出了制造业web服务合成语义描述的一种方法。     

Watermarking of digital geospatial datasets: a review of technical,legal and copyright issues

Gathering data accounts for more than 80% of the cost of any GIS project. Fast Internet connections and digital datasets threaten the investments of data producers through data piracy. Outside the GIS community this problem has been known for a long time, and possible solutions exist for digital imagery, formatted text, 3D meshes and so on, showing possible links to typical spatial data. This is largely achieved through embedding hidden information in a dataset without producing perceptible changes in the data, a process known as watermarking. The producer can recover the embedded information on request in order to produce evidence of ownership in a court, so the overall strategy relies on a legal basis rather than technical ones. This paper analyses the state-of-the-art for watermarking protection in digital geographical datasets. Digital imagery is demonstrably a more mature area than geographical information, even with multiple commercial vendors offering watermarking protection. 2D vector and point datasets have received less attention from the research community; however, 3D meshes have been considered by the CAD community and a handful of techniques are available for that case, and they are reviewed here.     

Directed movements in probabilistic time geography

This article studies probabilistic time geography for space–time prisms, that is, for situations where observers know the location of an agent at one time and then again at another time. In the intervening period, the agent would have moved freely, according to its time budget. The article demonstrates that the probability of finding the agent somewhere in the space–time prism is not equally distributed, so any attempt of a quantitative time geographic analysis must consider the actual probability distribution. This article develops, implements, and demonstrates this distribution. A preceding article introduced probabilistic time geography for space–time cones. With cones and prisms, the elementary space–time volumes of time geography are provided.     

Cardinal directions: a comparison of direction relation matrix and objects interaction matrix

How to express and reason with cardinal directions between extended objects such as lines and regions is an important problem in qualitative spatial reasoning (QSR), a common subfield of geographical information science and Artificial Intelligence (AI). The direction relation matrix (DRM) model, proposed by Goyal and Egenhofer in 1997, is one very expressive relation model for this purpose. Unlike many other relation models in QSR, the set-theoretic converse of a DRM relation is not necessarily representable in DRM. Schneider et al. regard this as a serious shortcoming and propose, in their work published in ACM TODS (2012), the objects interaction matrix (OIM) model for modelling cardinal directions between complex regions. OIM is also a tiling-based model that consists of two phases: the tiling phase and the interpretation phase. Although it was claimed that OIM is a novel concept, we show that it is not so different from DRM if we represent the cardinal direction of two regions a and b by both the DRM of a to b and that of b to a. Under this natural assumption, we give methods for computing DRMs from OIMs and vice versa, and show that OIM is almost the same as DRM in the tiling phase, and becomes less precise after interpretation. Furthermore, exploiting the similarity between the two models, we prove that the consistency of a complete basic OIM network can be decided in cubic time. This answers an open problem raised by Schneider et al. regarding efficient algorithms for reasoning with OIM.     

Fuzzy ontology-based foreshore identification from digital terrain models and very high-resolution airborne imagery through GEOBIA multi-scale analysis

ABSTRACT

Coastal mapping provides information for coastal zone protection. Land foreshore-border mapping is important, since it is related to land and property management, raising the need for an automated foreshore delineation framework. This research develops a multi-scale GEographic Object-Based Image Analysis method for foreshore identification from single-date very high-resolution imagery and DTM. Foreshore interpretation criteria were provided by the Greek Cadastral Office; however, they defined the foreshore-borders in an ill-defined manner. Thus, an ontology was designed and implemented to formalize the implicit spectral, geometric, and spatial relationships described in the interpretation criteria, and employ them during identification. The final ontology was delivered after three design phases: specification, conceptualization, and knowledge formalization. Imagery preprocessing through anisotropic morphological levelings reduced undesired spatial detail and noise. A three-level segmentation and ontological classification approach was developed to compute proper objects and assign them into semantic categories. Class definitions were created through fuzzy restrictions, representing value ranges of the employed properties. Evaluation of the extracted foreshore area with human-interpreted reference data showed a 77.9% quality, which was satisfactory. Evaluation by comparing the extracted foreshore border to the reference data showed an average deviation of 2.4 m, which is within Greek Cadastral Office specifications.     

基于语义类型本体的地名组配查询

地名查询方式多种多样,但它们都没有考虑地名语义类型因素,而语义类型恰恰是地名信息中相当重要的一部分。以地名语义分类为基础,构建了地名语义类型本体模型,并以郑州市为例进行了地名组配查询实验。结果表明,该方式对于检索地名的相关信息具有较好的辅助作用,为地名查询与检索作了有益的补充。     

Distributed geospatial information processing: sharing distributed geospatial resources to support Digital Earth

Abstract

This paper introduces a new concept, distributed geospatial information processing (DGIP), which refers to the process of geospatial information residing on computers geographically dispersed and connected through computer networks, and the contribution of DGIP to Digital Earth (DE). The DGIP plays a critical role in integrating the widely distributed geospatial resources to support the DE envisioned to utilise a wide variety of information. This paper addresses this role from three different aspects: 1) sharing Earth data, information, and services through geospatial interoperability supported by standardisation of contents and interfaces; 2) sharing computing and software resources through a GeoCyberinfrastructure supported by DGIP middleware; and 3) sharing knowledge within and across domains through ontology and semantic searches. Observing the long-term process for the research and development of an operational DE, we discuss and expect some practical contributions of the DGIP to the DE.     

Assessing the Effectiveness and Efficiency of Map Colour for Colour Impairments Using an Eye-tracking Approach

Colour impairments influences access to geographical information which is usually represented by colour maps. Three dimensions of colour: Hue, Saturation and Value (HSV), are intuitive and most critical visual variables in map design. In this paper, we specifically focus on colour deficiency of red-green colour impairments. A controlled experiment was designed and conducted to explore how three colour dimensions (HSV) affect the abilities of people with normal colour vision or with red-green colour impairments to distinguish colours in maps. An eye-tracking approach was applied to quantify the accuracy and response time by capturing user eye movements to analyse the effectiveness and efficiency. In this study, we used one section of the administrative map of Hebei Province to test participant responses to area features. Differences of effectiveness and efficiency across normal colour vision and red-green colour impairments were compared. Multiple comparisons among Hue, Saturation and Value were analysed. Results show that for both normal colour vision and red-green colour impairments, Hue is the most differentiable than Saturation and Value. Saturation and Value are at the same level to be differentiated and more difficult to be distinguished. Guidelines of designing maps for both normal colour vision and red-green colour impairments were derived. The results of this study can be helpful to improve the map designs for colour deficiency.