Trajectory compression algorithms eliminate redundant information in the history of a moving object. Such compression enables efficient transmission, storage, and processing of trajectory data. Although a number of compression algorithms have been proposed in the literature, no common benchmarking platform for evaluating their effectiveness exists. This paper presents a benchmarking framework for efficiently, conveniently, and accurately comparing trajectory compression algorithms. This framework supports various compression algorithms and metrics defined in the literature, as well as three synthetic trajectory generators that have different trade-offs. It also has a highly extensible architecture that facilitates the incorporation of new compression algorithms, evaluation metrics, and trajectory data generators. This paper provides a comprehensive overview of trajectory compression algorithms, evaluation metrics and data generators in conjunction with detailed discussions on their unique benefits and relevant application scenarios. Furthermore, this paper describes challenges that arise in the design and implementation of the above framework and our approaches to tackling these challenges. Finally, this paper presents evaluation results that demonstrate the utility of the benchmarking framework.
Real-time quality control (QC) of streaming natural resource data is needed to support the delivery of high quality data to system users. QC processes need to enable the identification of aberrations, as well as trends that may indicate degradation or component failures. These QC processes form a framework to support the goal of verified data delivered in a timely manner. In this paper, we investigate a method of computing Local Correlation Score (LCS) to detect anomalous patterns among sensor platforms in a concurrent manner. We use the R programming language and OpenMPI. Using empirical tests, we determine the benefits of computing the LCS in parallel, and on various sizes of clusters. We also analyze its use for real time mapping of Intelligent River data. Our results show that the LCS computed concurrently is an effective means for prompt quality assurance of natural resource data.
In recent years, several mobile devices with excellent performances have become accessible to people at affordable prices. The availability of this equipment, especially in the mobile sector, has encouraged research and development of increasingly complex applications ("Apps") for the visualization of large-scale scenes. However, 3D maps typically available through mobile version of so-called "spinning globes" do not allow the use of high definition data, due to their hardware limitations compared to desktop devices. As a result it often happens that a final user is navigating a real life familiar area, without being able to recognize its orography or specific features that are typical of the real world due to the poor resolution of the underlying 3D geometry. This is particularly amplified within mountain areas where crests, ridges and valleys are not adequately represented, due to the low resolution of the underlying digital terrain model, severely limiting the user's experience. The main contribution of the work presented by this paper is the enhanced user experience through high fidelity terrain representation, which is provided by an App that addresses the two aforementioned items.
In this work we describe our approach to efficiently create, handle and organize large-scale Structure-from-Motion reconstructions of urban environments. For acquiring vast amounts of data, we use a Point Grey Ladybug 3 omni directional camera and a custom backpack system with a differential GPS sensor. Sparse point cloud reconstructions are generated and aligned with respect to the world in an offline process. Finally, all the data is stored in a geospatial database. We incorporate additional data from multiple crowd-sourced databases, such as maps from OpenStreetMap or images from Flickr or Instagram. We discuss how our system could be used in potential application scenarios from the area of Augmented Reality.
Segmenting road regions from high resolution aerial images is an important yet challenging task due to large variations on road surfaces. This paper presents a simple and effective method that accurately segments road regions with a weak supervision provided by road vector data, which are publicly available. The method is based on the observation that in aerial images road edges tend to have more visible boundaries parallel to road vectors. A factorization-based segmentation algorithm is applied to an image, which accurately localize boundaries for both texture and nontexture regions. We analyze the spatial distribution of boundary pixels with respect to the road vector, and identify the road edge that separates roads from adjacent areas based on the distribution peaks. The proposed method achieves on average 90% recall and 79% precision on large aerial images covering various types of roads.
This paper presents a system for the creation of georeferenced 3D maps projectively textured with visual data gathered with an underwater robot. Using optical stereo cameras, maps are reconstructed and textured. Within the paper we propose a novel out-of-core 2D texture blending process that allows for high resolution texturing of complex 3D structure. Through the use of state-of-the-art model parameterization and texture atlasing the distortion of the final result can be minimized while the resolution of the original source imagery is maintained. We demonstrate both synthetic and real texturing results on 3D maps gathered with the Sirius Autonomous underwater vehicle (AUV). We discuss the implications for insufficient resolution when observing benthic features. Finally we conclude and discuss future directions for underwater 3D blending.
Today a huge amount of geospatial data is being created, collected and used more than ever before. The ever increasing observations and measurements of geo-sensor networks, satellite imageries, point clouds from laser scanning, geospatial data of Location Based Services (LBS) and location-based social networks has become a serious challenge for data management and analysis systems. Traditionally, Relational Database Management Systems (RDBMS) were used to manage and to some extent analyze the geospatial data. Nowadays these systems can be used in many scenarios but there are some situations when using these systems may not provide the required efficiency and effectiveness. More specifically when the geospatial data has high volume, high frequency of change (in both data content and data structure) and variety of structures, the conventional data storage systems cannot provide needed efficiency in online systems in terms of performance and scalability.In these situations, NoSQL solutions can provide the efficiency necessary for applications using geospatial data. This paper provides an overview of the characteristics of geospatial big data, possible solutions for managing and processing them. Then the paper provides an overview of the major types of NoSQL solutions, their advantages and disadvantages and the challenges they present in managing geospatial big data. Then the paper elaborates on serving geospatial data using standard geospatial web services with a NoSQL XML database as a backend.
Nowadays, we are witnessing formation of a new technological marvel: Internet of Things. This construction is able to combine in a particular operational entity all the bits and pieces of the world around us. Thus, why could not this unique establishment present the long-sought essence in the Nature of Things? The two pillars of modern fundamental science-relativity and quantum mechanics-are just approximate descriptions of some properties of such a constructive possibility. The machinery of the physical world develops on a cellular automaton model employing as the transformation rule the mechanism of distributed mutual synchronization with the property of fault-tolerance. This infrastructure yields traveling wave solutions that exactly correspond to the spectrum of the stable elementary particles of matter with an upper bound on the propagation speed. On top of the considered cellular automaton infrastructure there appears a secondary formation that constitutes the mechanism of the Holographic Universe that is the basis for the Internet of Things. The holographic activities determine all the quantum mechanics properties of the physical world including the nonlocality entanglement. For living systems the arrangement of the Internet of Things elucidates the most puzzling biological capability of morphogenesis that otherwise cannot find any reasonable explanation. In this paper, we present the world view of internet of things and the application of this methodology from geospatial computing to physics. We give specific details on applying IoT concept to geospatial analysis in various fields from agriculture to medicine. We also provide detailed analysis of the profound impact of internet of things on our physical world which is a vital knowledge when it comes to geospatial research. We present calendar variation of quantum world which can be used for geospatial data gathering by fine tuning the equipment based on the time of the year.
Image based localization is an important problem with many applications. In our previous work, we presented a two step pipeline for performing image based localization of mobile devices in outdoor environments. In the first step, a query image is matched against a georeferenced 3D image database to retrieve the "closest" image. In the second step, the pose of the query image is recovered with respect to the "closest" image using cell phone sensors. As such, a key ingredient of our outdoor image based localization is a 3D georeferenced image database. In this paper, we extend this approach to indoors by utilizing a 3D locally referenced image database generated by an ambulatory depth acquisition backpack that is originally developed for 3D modeling of indoor environments. We demonstrate retrieval rate of 94% over a set of 83 query images taken in an indoor shopping center and characterize pose recovery accuracy of the same set.
This project developed a new application platform software system to support automatically ITS facility's location decision-making system. The system, called PEDS (Platform for displacement Expert Design System), is in harmony with GIS, GPS and Artificial Intelligence Software to make design the position of ITS road facilities, automatically. We define the reference architecture for merging the heterogeneous technology such as Geoserver for open GIS, Postgre for open GIS DataBase and Drools for open expert software. We then implement a prototype model for developing ITS facility's location decision system. Finally, we discuss critical issues related to developing and operating a PEDS in an open source environment.
To increase performance, processor manufacturers extract parallelism through shrinking transistors and adding more of them to single-core chips and create multi-core systems. Although microprocessors performance continues to grow at an exponential rate, this approach generates too much heat and consumes too much power. These architectures not only introduce several complications but require tremendous efforts for organization of special software for parallel processing. In many cases, these difficulties are insurmountable. The programmers have to write complex code to prioritize the tasks or perform the task in parallel like extracting parallelism through threads in GPUs. One of the key issues for the programmers is how to divide the tasks in to sub-tasks. A faulty calculation may lead to increased data dependency which will slow the processor. Processor that performs more parallel operations can simultaneously increase the queuing delays. In both of the scenarios mentioned above, the relative cost of communication (also known as data transportation energy) between processing elements in microprocessor (or objects in parallel programming) is increasing relative to that of computation. This trend is resulting in larger caches for every new processor generation and more complex and costly latency tolerant mechanisms. Here we introduce a combinatorial architecture that has a unique property-multi-core running on a sequential code. This architecture can be used for both CPUs and GPUs. Some minor adjustments to a regular compiler are needed for loading. Especially, current mobile GPUs technologies are still relatively immature and require substantial improvements to enable wireless devices to perform the complex graphics-related functions. Our new architecture is more suitable for mobile GPUs/CPUs, i.e., mobile heterogeneous computing, with limited resources and relative greater performance.
With rapid increase of scope, coverage and volume of geographic datasets, knowledge discovery from spatial data have drawn a lot of research interest for last few decades. Traditional analytical techniques cannot easily discover new, implicit patterns, and relationships that are hidden into geographic datasets. The principle of this work is to evaluate the performance of traditional and spatial data mining techniques for analysing spatial certainty, such as spatial autocorrelation. Analysis is done by classification technique, i.e. a Decision Tree (DT) based approach on a spatial diversity coefficient. ID3 (Iterative Dichotomiser 3) algorithm is used for building the conventional and spatial decision trees. A synthetically generated spatial accident dataset and real accident dataset are used for this purpose. The spatial DT (SDT) is found to be more significant in spatial decision making.
Public space is one of the most important indicators of the quality of urban life. From a perspective of traditional urban planning, public spaces such as square, street, and plaza are significant elements of city environment to be designed. Designing public space in contemporary practice is rather a complicated decision-making process among diverse role players. It penetrates multiple layers of interest, ownership, and governance. This study stems from the crucial necessity of the common ground for understanding spatiotemporal context of public space. Since rapid urbanization increases complexity of the urban scene, the history of the place became difficult to be interpreted on site. Moreover, it is a quite compelling issue to determine cohesive spatial configuration across public and private spaces in urban design process. In this poster, we illustrated a series of experiments on data modeling and user interface design for four-dimensional media of urban design. Referring to data schema of empirically accessible data systems, we defined basic components that consist of public space including indoor public space in the private building, privately owned public open space, using Building Information Model (BIM) standard and geospatial Application Programming Interface (API) standard. The experiment delivers the scenario of dynamic timelines that conveys diverse user interaction over such physical description of urban space. The fundamental goal of this study is to develop an ontology of public space for mutually comprehensible design process among diverse role-takers over time.
Summary form only given. Geospatial data that exhibit time varying patterns are being captured faster than we are able to process them. We thus need machines to assist us in these tasks. One such problem is the automatic understanding of the behavior of moving objects for finding higher level information such as goals, intention etc. We propose a system that can solve one part of this complex task: automatic classification of movement patterns made by objects. In addition our system makes some simplifying assumptions: a) the object can be approximated as a moving point object (MPO) b) we consider interaction of a single MPO such as a car or mobile human, with static elements such as road networks and buildings e.g. airports, bus stops etc. on a terrain c) interactions between multiple MPOs are not considered. We use supervised machine learning algorithms to train the proposed system in classifying various patterns of spatiotemporal data. Algorithms such as Support Vector Machines and Decision Tree learning are trained with human labeled feature vectors that mathematically summarize how an MPO interacts with a landmark over time. Our feature vector incorporates a variety of geometric and temporal measurements such as the variable distances of the MPO to different points on the landmark, rate of change with time of variables such as distances and angles that are formed by the MPO with respect to the landmark. Simulated data created through graphical user interaction and agent-based modeling techniques are used to simulate MPO patterns over a representation of a real-world road network. The open source agent-based modeling tool Net Logo along with its GIS extension, and also the Agent Analyst module of ArcGIS are used to simulate large data sets. As future extensions, we are working on classification and prediction problems that involve multiple MPOs and landmarks.
Rainfall data is often collected by measuring the amount of precipitation collected in a physical container at a site. Such methods provide precise data for those sites, but are limited in granularity to the number and placement of collection devices. We use radar images of storm systems that are publicly available and provide rainfall estimates for large regions of the globe, but at the cost of loss of precision. We present a moving object database called Storm DB that stores decibel measurements of rain clouds as moving regions, i.e., we store a single rain cloud as a region that changes shape and position over time. Storm DB is a prototype system that answers rain amount queries over a user defined time duration for any point in the continental United States. In other words, a user can ask the database for the amount of rainfall that fell at any point in the US over a specified time window. Although this single query seems straightforward, it is complicated due to the expected size of the dataset: storm clouds are numerous, radar images are available in high resolution, and our system will collect data over a large timeframe, thus, we expect the number and size of moving regions representing storm clouds to be large. To implement our proposed query, we bring together the following concepts: (i) image processing to retrieve storm clouds from radar images, (ii) interpolation mechanisms to construct moving regions with infinite temporal resolution from region snapshots, (iii) transformations to compute exact point in moving polygon queries using 2-dimensional rather than 3-dimensional algorithms, (iv) GPU algorithms for massively parallel computation of the duration that a point lies inside a moving polygon, and (v) map/reduce algorithms to provide scalability. The resulting prototype lays the groundwork for building big data solutions for moving object databases.
Customizable tools that extend the functionality and enhance existing features within a software system are the keys to continued innovation. Depending on complexity, current and proposed projects tend to push the limits of existing functionality and require new tools to perform unique processes. Fortunately, software engineers and designers have taken this paradigm to heart and have created software systems with extensible architectures and frameworks. This paper presents one such customization for Esri ArcGIS that addresses the unique concerns and requirements of an ongoing project at the Center for Geospatial Information Technology involved with geo locating police reported vehicle crashes in the Commonwealth of Virginia. The tool takes advantage of theories and concepts from both computer science and geographic information systems to assist geocoders with evaluating, locating, and attributing crash data. Additionally, the tool provides a centralized web-based administrative portal for project managers.
In summer 2011 the US EPA's Climate Ready Estuaries program awarded funds to the Piscataqua Region Estuaries Partnership in coastal New Hampshire to further develop and use COAST (COastal Adaptation to Sea level rise Tool) for sea level rise adaptation planning. The New England Environmental Finance Center (EFC) worked with municipal staff, elected officials, and other stakeholders to select specific locations, vulnerable assets, and adaptation actions to model using COAST. The EFC then collected the appropriate base data layers, ran the COAST simulations, and provided visual, numeric, and presentation-based products in support of the planning processes underway in both locations. These products helped galvanize support for the adaptation planning efforts, and demonstrate utility of this new GIS based approach to community engagement for cost-benefit analysis of adaptations municipalities might take in response to sea level rise (SLR) and storm surge (SS).
We describe a framework for agent based modeling of moving point objects. Spatial movements are generated based on two overlapping ontologies. The first ontology includes the landmarks and descriptive outdoor behavior attributes. The second ontology includes indoor space and descriptive indoor behavior attributes. The modeling is based on ontology that includes the landmarks and descriptive behavior attributes of moving objects. The goal of our research in this area is to generate various spatial movements of point objects that can be classified into different known patterns. The agent behaviors can be modified semi-automatically based on changes in ontologies. This modeling is accomplished over a representation of a real-world data like road networks.
In previous geographic information inquiry, query condition is either fixed in program or providing an SQL inquiry mode for users. The former condition is unalterable while the latter demands users to be equipped with certain SQL query language knowledge. The article introduces how to use a rule engine to make inquiries through simple combination between natural semantic modules with the support of rule base. First, users formulate query plans through simple combination between natural language modules according to their own demands. Then, users deliver the query scheme to the rule engine for reasoning & matching, find the correct matching rule, and execute this rule. Finally, execution results are returned to users.
Digital Terrain Models (DTMs) are widely and intensively used as a computerized mapping and modeling infrastructure representing our environment. There exist many different types of wide-coverage DTMs generated by various acquisition and production techniques, which differ significantly in terms of geometric attributes and accuracy. In aspects of quality and accuracy most studies investigate relative accuracy relying solely on coordinate-based comparison approaches that ignore the local spatial discrepancies exist in the data. Our long-term goal aims at analyzing the absolute accuracy of such models based on hierarchical feature-based spatial registration, which relies on the represented topography and morphology, taking into account local spatial discrepancies exist. This registration is the preliminary stage of the quality analysis, where a relative DTM comparison is performed to determine the accuracy of the two models. This paper focuses on the second stage of the analysis applying the same mechanism on multiple DTMs to compute the absolute accuracy based on the fact that this solution system has a high level of redundancy. The suggested approach not only qualitatively computes posteriori absolute accuracies of DTMs, usually unknown, but also thoroughly analyzes the absolute accuracies of existing local trends. The methodology is carried out by developing an accuracy computation analysis using simultaneously multiple different independent wide-coverage DTMs that describe the same relief. A comparison mechanism is employed on DTM pairs using Least Squares Adjustment (LSA) process, in which absolute accuracies are computed based on theory of errors concepts. A simulation of four synthetic DTMs is presented and analyzed to validate the feasibility of the proposed approach.
