Design and implementation of an on-demand feature extraction web service to facilitate development of spatial data infrastructures

https://doi.org/10.1016/j.compenvurbsys.2008.07.002Get rights and content

Abstract

One of the main problems in urban and environmental management concerns the unavailability of reliable spatial data in a spatial data infrastructure (SDI) environment. The main reason for the problem of spatial data availability is the time-consuming nature of their manual production. The present paper proposes the development of on-demand data production Web services for the Internet using feature extraction techniques from satellite images as a solution to the problem. Such services allow users to connect to an on-demand data production Web service to produce the required data automatically if the users cannot find the required spatial data. In order to address and investigate this suggestion, a prototype system is developed. We have developed and implemented a system for automatic road extraction and describe it in special detail with a case study. Web service technologies and OGC (open geospatial consortium) frameworks are utilized for the development of the system to satisfy data and access interoperability in a SDI environment. The paper explains that the on-demand feature extraction Web service can facilitate the development of SDI by resolving the problem of spatial data availability. It also describes further research and different topics that should be considered in the development of SDIs to make such Web services operational.

Introduction

The development of spatial data infrastructure (SDI) has evolved as a central driving force in the management of spatial information over the last decade (Williamson, Grant, & Rajabifard, 2005). More than half the world’s countries claim that they are involved in some form of SDI development (Crompvoets, Rajabifard, Bregt, & Williamson, 2004). SDI concepts and models have also been recently used in different applications, such as disaster management (Mansourian, Rajabifard, Valadan Zoje, & Williamson, 2006), natural resource management, and wireless applications (Davies, 2003). With such wide range of activities, Masser (2005a) used the term ‘SDI phenomenon’ to describe the events that have taken place in this field over the last 10–15 years.

SDI is an initiative that intends to create an environment that will enable a wide variety of users to access, retrieve, and disseminate spatial data in an easy and secure way. In principle, SDIs allow the sharing of data; this is extremely useful, because it enables users like urban and environment managers to save resources, time, and energy when trying to acquire new datasets by avoiding duplication of expenses associated with the generation, maintenance, and integration of data. SDI aims to establish the relationship between people and data through appropriate policy-making, standardization activities, and the creation of accessing networks (the general SDI model: Rajabifard, Feeney, & Williamson, 2002). SDI is also an integrated, multi-level hierarchy of interconnected SDIs based on collaboration and partnerships among different stakeholders (SDI hierarchy: Rajabifard, Feeney, & Williamson, 2003).

Development of SDI is a matter of different technical, technological, social, institutional, and economical challenges (de Man, 2006, Mansourian et al., 2006, Masser, 2005b, Masser et al., 2007, Williamson et al., 2003). One of the important problems relates to the gap in spatial data (INSPIRE, 2003, Karatunga, 2002, Omran et al., 2006). The ‘gap’ in spatial data relates to unavailability of required data and/or the unreliability of available data due to low accuracy, being out-of-date, incompleteness, and similar issues. This problem is more critical in urban environments because of their fast development and changes. Since current manual methods for the production of spatial data is time-consuming and expensive, this issue is considered as an important problem for the development of SDI at its initial stages.

Remote sensing is a multidisciplinary science and technology that can collect data from any area in a short period of time. It offers a good solution that fulfills the data requirements of SDIs. Remote sensing offers data as an image in raster format, however, and most of the GIS (geographical information system) users would like to have spatial data at the feature level in a vector format. Automatic/semi-automatic feature extraction techniques from remote sensing data allow feature level data to be obtained with lower costs and shorter times than manual/traditional techniques.

This paper suggests the establishment of on-line Web services that provide users with data using an on-demand data production approach based on automatic/semi-automatic feature extraction techniques from satellite images. Web services equipped with automated/semi-automated feature extraction engines can thus be established. If the datasets required by a user are not available, then the user can connect to the Web service through the Internet, introduce the extent of the region of interest, and enter the required data layers to the engine. The system chooses the appropriate available satellite images from archive. Datasets that the user needs are produced automatically using advanced information processing techniques. The produced data can be added to the list of a spatial clearinghouse or catalogue service in order to make the data searchable and accessible for other users in the future.

Different challenges for making such a service operational and usable exist. From a SDI perspective, one technical challenge relates to spatial interoperability. Based on the OGC1 Reference Model,2 spatial interoperability refers to the “capability to communicate, execute programs, or transfer spatial data among various functional units in a manner that requires the user to have little or no knowledge of the unique characteristics of those units”. As this definition suggests the non-interoperability of spatial processing systems hampers the sharing of spatial data and services among software applications. In this context, two kinds of non-interoperability – data and access non-interoperability – can be identified3 (Peng, 2004).

Data non-interoperability implies that different spatial processing systems use internal data formats and produce data in formats that are different and in most cases proprietary. As a result, data sharing among different systems in a SDI environment is difficult. Access non-interoperability means that different spatial processing systems use proprietary software access methods with proprietary software interfaces, which restrict inter-process communication among various spatial processing systems. In other words, interface definition languages, communication protocols, communication ports, and even object transfer mechanisms vary in each software development platform. The software platform, which is used to develop the spatial processing system, thus imposes the use of specific and proprietary communication methods among various parts of the system. For this reason, different spatial processing systems developed by different software development platforms cannot communicate and share services automatically in an interoperable manner. This paper suggests an integrated utilization of Web service technologies (from the IT world) as well as OGC’s specifications and encodings (from the spatial information community) to resolve the non-interoperability problem for the proposed Web service.

In the context of this research, the development of a Web service that uses automatic feature extraction techniques to produce data for users in a SDI environment is ongoing. OGC frameworks and Web service technologies are used for the development of the system to provide spatial interoperability between the system and users. This paper describes the results of the project, including the development of a prototype system.

It is notable that current studies indicate that the deployment of spatial Web services is known as a significant factor facilitating the development of SDIs (Najar, Rajabifard, Williamson, & Giger, 2007). Web services can support the user in processing, accessing, and visualizing data. Most current activities developing spatial Web services and relevant standards focus on access (e.g. catalogue services) and visualization (e.g. Web Map Services), whereas less attention has been paid to processing aspects in the SDI environment. This research also addresses the benefits that can be gained from spatial Web services in the context of processing tasks within a SDI environment.

Section snippets

Methodology and framework

As highlighted earlier, automatic feature extraction techniques, OGC frameworks, and Web service technologies are three main components required in the development of a Web-based service that (1) has the capability of on-demand data production from remote sensing data and (2) acts as a step forward to facilitate SDI development. This section describes the algorithm, methodologies, and frameworks utilized for development of the prototype system.

Development of a prototype service: a case study

A prototype system is developed as part of this research in order to investigate and demonstrate the proposed Web service, which is an on-demand data production system from satellite images. The system is developed based on Web client–server architecture. An automatic road extraction engine, raster-to-vector conversion engine, GML data production engine, WMS, and database construct the major components of the system at the server’s side (Fig. 2). Proper user interfaces at the client’s side are

Conclusions and future trends

The gap in spatial data is one of the major issues for developing SDIs. This paper proposes a Web-based, on-demand data production service on the Internet that uses feature extraction techniques from satellite images as a solution to this problem.

In order to investigate this suggestion, a prototype service was developed. OGC frameworks and Web service technologies were utilized for the development of the system; using such standards, data and access interoperability were achieved. The results

References (45)

  • Agouris, P., Gyftakis, S., & Stefanidis, A. (2001). Quality-aware deformable models for change detection. In...
  • P. Agouris et al.

    Differential snakes for change detection in road segments

    Photogrammetric Engineering and Remote Sensing

    (2001)
  • C. Baillard et al.

    A plane sweep strategy for the 3D reconstruction of buildings from multiple images

    International Archives of Photogrammetry and Remote Sensing

    (2000)
  • J.C. Bezdek

    Pattern recognition with fuzzy objective function algorithms

    (1981)
  • S. Chatterjee et al.

    Developing enterprise Web services: An architect’s guide

    (2003)
  • T. Chen et al.

    A wavelet transform based method for road centerline extraction

    Photogrammetric Engineering and Remote Sensing

    (2004)
  • J. Crespo et al.

    The flat zone approach: A general low-level region merging segmentation method

    Signal Processing

    (1998)
  • J. Crompvoets et al.

    Assessing the world wide developments of national spatial data clearinghouses

    International Journal of Geographical Information Science

    (2004)
  • Davies, J. (2003). Expanding spatial data infrastructure model to support spatial wireless applications. PhD thesis,...
  • W.H.E. de Man

    Understanding SDI; complexity and institutionalization

    International Journal of Geographical Information Science

    (2006)
  • J.C. Dunn

    Some recent investigations of a new fuzzy partition algorithm and its application to pattern classification problems

    Journal of Cybernetics

    (1974)
  • X. Hu et al.

    A robust method for semi-automatic extraction of road centerlines using a piecewise parabolic model and least square template matching

    Photogrammetric Engineering and Remote Sensing

    (2004)
  • Cited by (11)

    • Introduction to GPS/GNSS technology

      2021, GPS and GNSS Technology in Geosciences
    • Intelligent services for discovery of complex geospatial features from remote sensing imagery

      2013, ISPRS Journal of Photogrammetry and Remote Sensing
      Citation Excerpt :

      Feature extraction from remote sensing imagery provides a time-efficient and cost-effective way to generate geospatial data in a vector format. Mansourian et al. (2008) present the design and implementation of a feature extraction Web Service. An automatic feature extraction algorithm is provided in the Web Service to allow online extraction of roads from satellite imagery.

    • Identifying orthoimages in Web Map Services

      2012, Computers and Geosciences
      Citation Excerpt :

      The accessibility to geospatial resources influences any geospatial-based research. New approaches, such as on-demand data production (Mansourian et al., 2008), assume discoverability of proper geospatial resources as an input. The discovery of resources within an SDI is based on the digital library paradigm (Béjar et al., 2009).

    • Improving data management and dissemination in web based information systems by semantic enrichment of descriptive data aspects

      2010, Computers and Geosciences
      Citation Excerpt :

      Information systems for sustainable and adaptive Integrated Water Resources Management (IWRM) depends on the availability and accessibility of informations for system analysis and modelling, social learning, and sustainable decision making (Flügel, 2007). Spatial data and related technologies such as geographical information systems (GIS), global positioning systems (GPS), or remote sensing provide information on abiotic, biotic and socioeconomic conditions, and are proven crucial for effective collaborative decision-making in the IWRM context (Kiehle, 2006; Flügel, 2007; Mansourian et al., 2008). Large amount of observational and modelling data with extensive geographical and temporal coverage were collected in the past (McGuire et al., 2008; Schindler and Diepenbroek, 2008), and data processing requirements will continue to increase due to the advance of new sensors and automatic data acquisition tools (Butenuth et al., 2007).

    • Downscaling on Demand: Examples in Forest Canopy Mapping

      2017, Integrating Scale in Remote Sensing and Gis
    View all citing articles on Scopus
    View full text