Automatic alignment of contemporary vector data and georeferenced historical maps using reinforcement learning

ABSTRACT

With large amounts of digital map archives becoming available, automatically extracting information from scanned historical maps is needed for many domains that require long-term historical geographic data. Convolutional Neural Networks (CNN) are powerful techniques that can be used for extracting locations of geographic features from scanned maps if sufficient representative training data are available. Existing spatial data can provide the approximate locations of corresponding geographic features in historical maps and thus be useful to annotate training data automatically. However, the feature representations, publication date, production scales, and spatial reference systems of contemporary vector data are typically very different from those of historical maps. Hence, such auxiliary data cannot be directly used for annotation of the precise locations of the features of interest in the scanned historical maps. This research introduces an automatic vector-to-raster alignment algorithm based on reinforcement learning to annotate precise locations of geographic features on scanned maps. This paper models the alignment problem using the reinforcement learning framework, which enables informed, efficient searches for matching features without pre-processing steps, such as extracting specific feature signatures (e.g. road intersections). The experimental results show that our algorithm can be applied to various features (roads, water lines, and railroads) and achieve high accuracy.     

Ontologies and representation spaces for sketch map interpretation

In this paper, we present a systematic approach to sketch map interpretation. The method decomposes the elements of a sketch map into a hierarchy of categories, from the material sketch map level to the non-material representational sketch map level, and then interprets the sketch map using the five formal representation spaces that we develop. These spaces (set, graph, metric and Euclidean) provide a tiered formal representation based on standard mathematical structures. We take the view that a sketch map bears information about the physical world and systematises this using extensions of existing formal ontologies. The motivation for this work is the partially automatic extraction and integration of information from sketch maps. We propose a set of ontologies and methods as a first step in the direction of a formalisation of partially automatic extraction and integration of sketch map content. We also see this work as a contribution to spatial cognition, where researchers externalise spatial knowledge using sketch mapping. The paper concludes by working through an example that demonstrates the sketch map interpretation at different levels using the underlying method.     

Does base map size and imagery matter in sketch mapping?

Sketch mapping has been an important data collection technique for geographers since the 1960s. Structured sketch mapping requires participants to draw spatial data onto a base map containing cartographic information, in order to assist with spatial referencing. A concern that base map characteristics may influence sketch map content has been raised repeatedly in the research literature. However little scholarly attention has been paid to systematically testing the effect of base maps.This paper aims to test the effects of base map size and imagery on structured sketch maps of avoidance behaviour in university students. Using an experimental design, 272 sketch maps were compared for differences in: sketch map style; the location of collective avoidance hotspots; the extent of the reported area avoided; the number of reported areas avoided; the intensity of avoidance; and the tortuosity of sketch map features.No significant differences were found between base maps in sketch map style or the size, intensity or number of areas avoided. Provision of larger base maps caused respondents to draw more detailed sketch maps. Collective avoidance hotspots shifted location slightly between base maps, probably due to difficulties interpreting aerial photographs.Sketch map content appears to be remarkably robust to changes in base map. Base maps appear to assist respondents with spatial referencing rather than cueing respondents to report specific features.     

Sketch Maps and Qualitative GIS: Using Cartographies of Individual Spatial Narratives in Geographic Research

Mental and sketch maps have a long tradition in modern geography. Little theoretical and methodological distinction has been made, however, between different hand mapping approaches. Mental maps emerged from behavioral geography of the spatial science tradition as a technique to understand human behaviors based on peoples’ perceptions of their spatial environment. More recently, sketch maps have been used in participatory and qualitative geographic information systems (QGIS) to develop cartographies of group and individual spatial narratives. They are a tool that helps achieve QGIS's objectives of collecting unique spatial data of individual experiences, visualizing socio-spatial processes, breaking down particular barriers of positionality in research, and developing new uses of GIS. Two case studies illustrate the use of sketch maps in research, including a project examining job accessibility among working poor individuals and a study of the experiences of fear and safety in public spaces among lesbian, gay, bisexual, and transgender community members. Sketch maps in QGIS have many methodological merits. They add an invaluable dimension to the qualitative interview process, offer countermapping perspectives, generate detailed spatial information of individuals, and facilitate data interpretation.     

Same space – different perspectives: comparative analysis of geographic context through sketch maps and spatial video geonarratives

ABSTRACT

The importance of including a contextual underpinning to the spatial analysis of social data is gaining traction in the spatial science community. The challenge, though, is how to capture these data in a rigorous manner that is translational. One method that has shown promise in achieving this aim is the spatial video geonarrative (SVG), and in this paper we pose questions that advance the science of geonarratives through a case study of criminal ex-offenders. Eleven ex-offenders provided sketch maps and SVGs identifying high-crime areas of their community. Wordmapper software was used to map and classify the SVG content; its spatial filter extension was used for hot spot mapping with statistical significance tested using Monte Carlo simulations. Then, each subject’s sketch map and SVG were compared. Results reveal that SVGs consistently produce finer spatial-scale data and more locations of relevance than the sketch maps. SVGs also provide explanation of spatial-temporal processes and causal mechanisms linked to specific places, which are not evident in the sketch maps. SVG can be a rigorous translational method for collecting data on the geographic context of many phenomena. Therefore, this paper makes an important advance in understanding how environmentally immersive methods contribute to the understanding of geographic context.     

Early Instruction in Geography: An Exploration in the Ecology of Kindergarten and First-Grade Geography Education

Abstract

Student drawn maps have been relatively neglected as an evaluation tool, despite their potential utility for assessing geographic knowledge. Comparison of students' sketch maps at the beginning and end of the year can inform the teacher about how students' representations of the world change over the course of instruction. Used as a pretest, they can be used to diagnose areas of strength and weakness within a class; as a posttest they can be used to assess students' achievement and the effectiveness of instruction. In this paper we provide an easy method for sorting and summarizing sketch map data, which is illustrated using maps collected from two sixth grade classes. We also outline a framework to guide teachers in generating questions to ask about their own students' maps.     

认知地图的空间句法研究

将意识层面的空间意象外化为可被度量及分析的对象,一直是行为学、地理学、旅游学等学科关注的重要课题。本文主要从构建空间意象的结构而非意义入手,对该领域研究中的两种重要方法,即认知地图法和空间句法进行了比较,意图找到两种方法在空间结构性认知方面的相关性。基于句法空间结构的表达模型,分别就长沙太平街历史街区集成度及其认知地图的调研结果进行了比对,通过真实地图句法集成度与579 份认知地图意象要素出现频率等指标的比较,发现集成度指标与意象出现频率等指标存在较好的相关性。在对认知地图样本进行轴线化转换并比照真实地图的轴线索引后,对真实地图句法集成度与579 份认知地图平均句法集成度进行比较,发现认知地图与真实地图的句法分析在结果上仍有较好的相关性,且节点、标志物两要素的出现频次等指标也与所在街道的集成度存在较为明显的对应性。结论说明空间构形对人们结构性意象的形成有着重要的影响力,认知地图的句法分析对揭示人们空间知识的形成机制具有一定的参考价值,两种方法有相互补充的可能性。     

基于跟踪调查的北京城市空间感知研究——以中国科学院大学硕士研究生为例

城市空间感知是行为地理学的经典研究课题,论文采用认知地图并辅以问卷调查方法,在一年时间内跟踪调查中国科学院大学雁栖湖校区2017级硕士研究生,根据手绘地图分析中国科学院大学(国科大)硕士研究生城市空间感知的过程与特征。结果表明：① 有别于凯文·林奇(K. Lynch)的空间意象5大要素,国科大研究生手绘地图的基本要素主要为标志物、节点、道路,其中标志物占主导地位;② 根据要素组合特征,可将认知地图划分为散点型、过渡型和拓扑型3类,散点型地图占据绝对数量优势,随驻留北京时间延长,认知地图结构渐趋复杂;③ 国科大研究生北京城市空间感知表现出偏好特殊性、空间非均衡性、个体差异性、感知过程不完整性等特点;④ 远距离感知、城市空间尺度大、受访者逗留时间短等主客观因素制约了北京城市整体可意象性。     

Integration of multi-source remote sensing data for land cover change detection

The objective of this study is to develop a methodology to integrate multi-source remote sensing data into a homogeneous time series of land cover maps in order to carry out change detection. We developed a method to increase the comparability between land cover maps coming from panchromatic aerial photographs and SPOT XS (multi-spectral) data by equalizing their levels of thematic content and spatial details. The methodology was based on the hypotheses that: (1) map generalization can improve the integration of data for change detection purpose, and (2) the spatial structure of a land cover map, as measured by a set of landscape metrics, is an indicator of the level of generalization of that map. Firstly, the methodology for data integration was developed by using land cover maps generated from near-synchronous data. Results revealed that, by controlling successively the parameters that influence the level of map generalization, the percentage of agreement between the near-synchronous land cover maps can be increased from 42% to 93%. The computation of five landscape metrics for a set of generalized land cover maps and for the target map allowed us to optimize the level of generalization by measuring the similarity in landscape pattern of the maps. The optimum level of generalization of the land cover map obtained from the aerial photographs for comparison with a land cover map derived from SPOT XS data was found at a resolution of 41m for two generalization levels of the thematic content. The spatial structure of a land cover map, as measured by a set of landscape metrics, is thus a good indicator of the level of generalization of this map. Secondly, the method was applied by integrating a land cover map obtained from aerial photographs of 1954 with a land cover map obtained from a SPOT XS image of 1992.     

Formal representation of qualitative direction

This paper reviews formal approaches to representing spatial knowledge about qualitative direction. Unlike geometric direction information, qualitative information does not employ numerical values but relies on comparison. The qualitative approach is often regarded as suitable for capturing commonsense concepts and thus is relevant to human-centered interfaces for spatial information systems. To establish a context for the work on qualitative direction, we preset a brief history of the development of qualitative temporal and spatial representations from different scientific perspectives. We identify main focal areas of these representations of spatial direction and propose a taxonomy. In the light of more than three decades of fruitful research, we obtain a map of formal representations that reveal interrelationships between different research strands in the field.     

A Commentary on the Use of Touch for Accessing On‐Screen Spatial Representations: The Process of Experiencing Haptic Maps and Graphics

The growth of the Internet and the digital revolution have meant increased reliance on electronic representations of information. Geospatial information has been readily adapted to the world of cyberspace, and most Web pages incorporate graphics, images, or maps to represent spatial and spatialized data. But flat computer screens do not facilitate a map or graph experience by those who are visually impaired. The traditional method for compensating for nonvisual access to maps and graphics has been to construct hard‐copy tactile maps. In this article, we examine an electronic accommodation for nonvisual users—the haptic map. Using new and off‐the‐shelf hardware—force feedback and vibrotactile mice—we explore how touch can be combined with virtual representations of shapes and patterns to enable nonvisual access to onscreen map or graphic material.     

World Patterns of Geographic Literacy Based on Sketch Map Quality

This paper discusses world patterns of geographic literacy and illiteracy based on the quality of sketch maps of the world drawn by university students in first-year geography classes. The study, sponsored by the International Geographical Union and funded by the National Geographic Society, contains 3,568 sketch maps of the world collected from 75 sites in 52 countries. For the purposes of this paper, geographic literacy is equated with the number of nations and continents included on the sketch maps. The individual map scores are aggregated to determine the average number of items per map for each site and nation. Our set of maps provides a baseline of world variations in geographic literacy just prior to the end of the Cold War. The sketch map techniques can be used to monitor changes in geographic literacy.     

Comparison of thematic maps using symbolic entropy

Comparison of thematic maps is an important task in a number of disciplines. Map comparison has traditionally been conducted using cell-by-cell agreement indicators. More recently, other methods have been proposed that take into account not only spatially coincident cells in two maps, but also their surroundings or the spatial structure of their differences. The objective of this article is to propose a framework for map comparison that considers (1) the patterns of spatial association in two maps, in other words, the map elements in their surroundings; (2) the equivalence of those patterns; and (3) the independence of patterns between maps. Two new statistics for the spatial analysis of qualitative data are introduced that are based on the symbolic entropy of the maps. As well, all inferential elements to conduct hypothesis testing are developed. The framework is illustrated using real and simulated maps.     

SCALING TECHNIQUES IN THE ANALYSIS OF ENVIRONMENTAL COGNITION DATA *

The utility of nonmetric, multidimensional-scaling techniques is demonstrated for the analysis and collection of environmental-cognition data. By comparing the multidimensional-scaling solutions of a real-setting map to scaling solutions for sketch maps and two psychophysical, distance-scaling procedures, we demonstrate that magnitude estimation of actual interpoint distances is comparable in accuracy to sketch maps when produced without constraints, or when subjects are given a specified list of landmarks to include on their maps. Triadic comparisons of actual interpoint distances were less accurate than the three other techniques.     

SCALING TECHNIQUES IN THE ANALYSIS OF ENVIRONMENTAL COGNITION DATA∗

The utility of nonmetric, multidimensional-scaling techniques is demonstrated for the analysis and collection of environmental-cognition data. By comparing the multidimensional-scaling solutions of a real-setting map to scaling solutions for sketch maps and two psychophysical, distance-scaling procedures, we demonstrate that magnitude estimation of actual interpoint distances is comparable in accuracy to sketch maps when produced without constraints, or when subjects are given a specified list of landmarks to include on their maps. Triadic comparisons of actual interpoint distances were less accurate than the three other techniques.     

Curvedness feature constrained map matching for low-frequency probe vehicle data

Map matching method is a fundamental preprocessing technique for massive probe vehicle data. Various transportation applications need map matching methods to provide highly accurate and stable results. However, most current map matching approaches employ elementary geometric or topological measures, which may not be sufficient to encode the characteristic of realistic driving paths, leading to inefficiency and inaccuracy, especially in complex road networks. To address these issues, this article presents a novel map matching method, based on the measure of curvedness of Global Positioning System (GPS) trajectories. The curvature integral, which measures the curvedness feature of GPS trajectories, is considered to be one of the major matching characteristics that constrain pairwise matching between the two adjacent GPS track points. In this article, we propose the definition of the curvature integral in the context of map matching, and develop a novel accurate map matching algorithm based on the curvedness feature. Using real-world probe vehicles data, we show that the curvedness feature (CURF) constrained map matching method outperforms two classical methods for accuracy and stability under complicated road environments.     

定性地理信息系统在城市社会空间研究中的应用

定性地理信息系统(定性GIS)是西方人文地理学定性研究的重要方法。它集成了定性和空间元素,更全面地阐明空间背景是如何被社会建构的。这种多元要素的整合,可以表达人们生活空间的多样性和复杂性,有助于展示生活路径的三维可视化细节,实现可视化社会空间过程。本文试图梳理西方定性GIS的方法论并探讨其在基于情感表达的个体数据采集分析、基于混合方法的可视化活动空间研究以及基于多重互动视角的社会空间研究中的应用。定性GIS包含微观和宏观层面的多重含义,使得研究者可以从多个有利位置观察问题并参与其中,有助于城市社会空间研究理论创新,为社会空间研究提供可探索空间。最后,从学科建设、方法论拓展、研究成果表达和研究议题选择等方面对中国开展同类研究进行展望。     

Porphyry Copper Deposit Tract Definition – A Global Analysis Comparing Geologic Map Scales

Geologic maps are a fundamental data source used to define mineral-resource potential tracts for the first step of a mineral resource assessment. Further, it is generally believed that the scale of the geologic map is a critical consideration. Previously published research has demonstrated that the U.S. Geological Survey porphyry tracts identified for the United States, which are based on 1:500,000-scale geology and larger scale data and published at 1:1,000,000 scale, can be approximated using a more generalized 1:2,500,000-scale geologic map. Comparison of the USGS porphyry tracts for the United States with weights-of-evidence models made using a 1:10,000,000-scale geologic map, which was made for petroleum applications, and a 1:35,000,000-scale geologic map, which was created as context for the distribution of porphyry deposits, demonstrates that, again, the USGS US porphyry tracts identified are similar to tracts defined on features from these small scale maps. In fact, the results using the 1:35,000,000-scale map show a slightly higher correlation with the USGS US tract definition, probably because the conceptual context for this small-scale map is more appropriate for porphyry tract definition than either of the other maps. This finding demonstrates that geologic maps are conceptual maps. The map information shown in each map is selected and generalized for the map to display the concepts deemed important for the map maker’s purpose. Some geologic maps of small scale prove to be useful for regional mineral-resource tract definition, despite the decrease in spatial accuracy with decreasing scale. The utility of a particular geologic map for a particular application is critically dependent on the alignment of the intention of the map maker with the application.