research article a framework for sharing and integrating remote sensing and gis...

Research ArticleA Framework for Sharing and Integrating Remote Sensing andGIS Models Based on Web Service

Zeqiang Chen123 Hui Lin134 Min Chen13 Deer Liu15

Ying Bao16 and Yulin Ding17

1 Institute of Space and Earth Information Science The Chinese University of Hong Kong Shatin Hong Kong2 Collaborative Innovation Center for Geospatial Information Technology Wuhan University Wuhan 430079 China3 Shenzhen Research Institute The Chinese University of Hong Kong Shenzhen 518057 China4Department of Geography and Resource Management The Chinese University of Hong Kong Shatin Hong Kong5 Faculty of Architectural and Survey Engineering Jiangxi University of Science and Technology Ganzhou 341000 China6 International Institute for Earth System Science Nanjing University Nanjing 210023 China7 State Key Laboratory of Information Engineering in Surveying Mapping and Remote SensingWuhan University Wuhan 430079 China

Correspondence should be addressed to Hui Lin huilincuhkeduhk

Received 11 January 2014 Accepted 27 February 2014 Published 11 May 2014

Academic Editors P Bala P Ji and S Xiang

Copyright copy 2014 Zeqiang Chen et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Sharing and integrating Remote Sensing (RS) and Geographic Information SystemScience (GIS) models are critical for developingpractical application systems Facilitating model sharing andmodel integration is a problem for model publishers andmodel usersrespectively To address this problem a framework based on a Web service for sharing and integrating RS and GIS models isproposed in this paper The fundamental idea of the framework is to publish heterogeneous RS and GIS models into standardWebservices for sharing and interoperation and then to integrate the RS and GIS models using Web services For the former a ldquoblackboxrdquo and a visual method are employed to facilitate the publishing of the models as Web services For the latter model integrationbased on the geospatial workflow and semantic supported marching method is introduced Under this framework model sharingand integration is applied for developing the Pearl River Delta water environment monitoring system The results show that theframework can facilitate model sharing and model integration for model publishers and model users

1 Introduction

Sharing and integrating Remote Sensing (RS) and Geo-graphic Information SystemScience (GIS)models are impor-tant for solving comprehensive complex and multidisci-plinary problems such as urban growth [1] and environmen-tal applications [2 3] However with the inherent naturalcharacteristics such as heterogeneity (eg structure andexecution environment) and dispersion (eg owner andrunning server) and the man-made characteristics such asprivacy (eg authorization-required and classified) of RSand GIS models simple and effective sharing and integratingRS and GISmodels are still challenges althoughmany efforts

have been made to improve model sharing and integrationstrategies [4ndash6]

Recent studies have focused on model sharing and inte-gration problems [7 8] Decades ago a number of modelrepresentation approaches and languages such as alge-braic modeling languages [9ndash11] logic-based systems [12]relational-based approaches [13 14] graph-based languages[4 15ndash18] and structured modeling [19 20] were proposedto handle model sharing and integration issues by providinga consistent representation method but they have limitedsupport for model sharing in a distributed environment[21] Distributed object technologies have been employed toaccess remote objects in a distributed environment such as

Hindawi Publishing Corporatione Scientific World JournalVolume 2014 Article ID 354919 13 pageshttpdxdoiorg1011552014354919

2 The Scientific World Journal

the Common Object Request Broker Architecture (CORBA)[22] the Distributed Component Object Model (DCOM)[23] and the Java Remote Method Invocation (Java RMI)[24] the direct use of these three technologies by pro-grammers is becoming less common for their shortcomingstightly coupled protocols are effective for building a specificapplication but are not sufficiently flexible and protocols areconstrained by their vendor implementations platforms andlanguages [25]

In recent years standard technologies specifications andframeworks have been developed for model sharing andintegration examples include theModel Driven Architecture(MDA) and the open modeling interface (OpenMI) as wellas several modeling frameworks MDA is a software designapproach developed by the Object Management Group toprovide an open and vendor-neutral approach for softwaresystems using specifications such as The Unified ModelingLanguage the Meta Object Facility the XML MetadataInterchange and the Common Warehouse Meta-model thatbuilt coherent schemes for authoring publishing and man-aging models with the platform independent model PIMand the platform-specific model PSM [26 27] The MDA ispromising but the PIM and the PSM do not include standardspecification mapping which complicates application of theMDA The open modeling interface (OpenMI) is a softwarecomponent interface that allows time-dependent modelsto share and exchange information at runtime makingmodel integration feasible at the operational level OpenMIemphasizes interoperability between otherwise independentmodels [28] Dozens of modeling and reuse frameworkssuch as the spatial modelling environment (SME) [29 30]and the interactive component modelling system (ICMS)[31] have been developed to handle model sharing andintegration problems but most of them are highly field-related platform-related and internet-inaccessible whichlimits their application [5]

With the development of information technology (IT)Web service and service-oriented architecture (SOA) aremore commonly used in model sharing and model integra-tion [2 32ndash34] Some advantages include sharing models asworldwide services applied to various popular applications[35] and integrating data and models from anywhere in theworld [36] In addition with the advent of the advancednotion of ldquoModel as a Servicerdquo [37ndash40] the Web servicemethod creates new opportunities for sharing and integratingmodels including GIS and RS models

The objective of this paper is to design a Web service-based method to share and integrate RS and GIS modelsThemain idea of model sharing and integration based on Webservice is to release various heterogeneousRS andGISmodelsas Web services for model sharing and then to integrate thehomogeneousWeb services for model integration Followingthis idea this paper publishes the methods model and themodel integration as Web services and applies this to thewater environmentmonitoring in the Pearl RiverDelta regionin China

The organization of this paper is as follows Section 2addresses the main idea of web service-based model shar-ing and integration and the proposed method Section 3

describes the implementation of the proposed method forthe PRD water environment monitoring system FinallySection 4 concludes this paper and outlines future work

2 Methodologies

A model is a representation of one or more processes thatare believed to occur in the real word [41] From a userrsquosperspective a model can be formalized as a tuple as 119872 =ID 119873 119868 119874 IMPLMD where119872 is the model ID presentsthe unique identification of themodel119873 is its name and 119868119874IMPL andMD are the input the output the implementation(executable programs) and the metadata of the modelrespectively

Web service offers a way to share and interoperatemodels[40] The method is based on Simple Object Access Protocol(SOAP) and Web Services Description Language (WSDL)WSDL provides a function-centric description of the Webservices including their inputs outputs and exception hand-ing [42] which map to the structure of the formalized modeltuple Because the method is standardized and loosely cou-pled [40] heterogeneous RS andGISmodels become ldquohomo-geneousrdquo Web services after they are published Integratingheterogeneous RS andGISmodels is equivalent to integratingthe ldquohomogeneousrdquo Web services because the models arepublished as Web services Web service composition is amethod of integrating Web services [33 34 43] Web servicecomposition aggregates multiple services into a single newservice for a certain functionality that a single primary servicecannot provide [44] If a model that runs on a desktop orlocal area network environment is called the local model anda model that is published with Web service is called a Webmodel the aim of sharing and integrating the GIS and RSmodels is to publish the local model to a Web model andthen combine theWebmodelsThe framework of sharing andintegratingGIS andRSmodels based onWeb service is shownin Figure 1

21 Model Sharing with Web Service Model sharing is thepublishing of local models as Web models using a Webservice Developing a Web service is time intensive andrequires specific professional skills In addition many GISand RS model experts aim to concentrate their efforts ondeveloping more powerful models rather than on improvingtheir programming skills Considering these factors a blackbox approach combined with a visual method is introducedto facilitate the publishing of local models to Web services Ablack box [45] is a method that can be used without knowinghow its inner algorithm works the user only needs to knowthe input and output characteristics The exposed entry ofthe black box is a visual user interface that is illustrated inFigure 2 A visual method for manipulating a model is aneffective way to reduce development time and training [46]The visual interface hides the cumbersome implementationof the Web service

For this method ldquoinsiderdquo the black box is criticalgenerating Web services from GIS and RS models withtheir names inputs outputs and executable programs The

The Scientific World Journal 3

GIS models

RS models

Localserver

Distributed models Web Web service Integratedservicemodels

Integratedservicemodels

Publishing models to Web services Models integrationbased on Web service

Heterogeneousmodels

Homogeneousservice

Procedure

Localserver

Localserver

SOAP

Web service

SOAP

Web service

SOAP

Web service

Figure 1 The framework of GIS and RS model sharing and integration based on Web service

RS

Black box

Web service

GIS model

model

Figure 2 The black box approach combined with a visual methodfor publishing local models to Web services

inputs to the black box include description parametersand the output includes the Web functions An interfacemethod is introduced to perform these functions Interface-based programming is a common programming methodin high-level programming languages An interface is areference-type object that defines methods without definingthe implementation It is a signature for interactingwith otherclasses or interfaces An interface-based method separates amethod definition and its implementation and makes codemore reusable robust revisable and abstract The purposesof adopting an interface-based method are (1) to providea uniform method of describing the inputs and outputsfor a model ldquooutsiderdquo the black box which facilitates theinteraction between a model and the black box and (2) tomake the Web service implementation easy with a unifiedinterface ldquoinsiderdquo the black box Two important items for

the interface-based method are the interface definition andthe interface implementation

For the interface definition the key of publishing amodel to aWeb service is mapping the description structuresbetween the model and the Web service The Operation ina WSDL document is similar to a method or function ina traditional programming language An executable modelis also similar to a method or function in a traditionalprogramming language The outline for mapping the rela-tionship between a model and a Web service is shown inFigure 3 The input output and execution code of a modelmap to the input output and execution code of a Webservice respectively The execution code of the model is itsprogramming implementation which is consistent with theWeb service

In Figure 2 the name input and output of amodelmap tothe name input and output of an interface respectively Theinput and output parameters are set with names and typesThe types are basic programming types such as string doubleand intThese parameter names and types are consistent withthose of the model executable program inputs and outputs

For interface implementation models deployed in alocal server may occur in one of many forms for exampleexecutable file (EXE) script dynamic link library (DLL)or other written program languages These form-executableprograms are called third-party programs Invoking a third-party program is essential to run a local model High-level programming languages are able to invoke third-partyprograms For instance Java uses runtime and process classes


Model

executioncode

ModelInput

Output

Web service

WSDL

Interfaceoutput operation (input)

Execution code

Output operation (input)

The mapping relationship between A and BA B

middot middot middot

Figure 3 Mapping a model to a Web service

to run third-party programs and in C DllImport is usedto execute third-party programs A Web service develop-ment library supports the development of Web servicesC includes Web service classes in the NET Framework aslibrary classes for the development of Web services Java alsohas library classes for Web service development Java hasmany open source projects such as Axis2 that can be used todevelop and deploy Web services In Figure 2 a Web serviceskeleton is shown for the interface programmingmethod thatmaps a model to a Web service When a concrete interface isset the third-party program fills in the skeleton to realize theWeb service

22 Model Integration Using a Web Service RS and GISmodels are homogeneous Web services after the models arepublished intoWeb models The model integration challengeis equivalent toWeb service integrationWeb service integra-tion consists of combining different but associated Web ser-vices on theWeb AWeb service demonstrates its capabilitiesbyWSDLwith operations ComposingWeb services includescombining operations using a logical process Web serviceintegration creates an order-logic combination of relatedWebservices

Web service composition has been studied for yearsIndustry and academia each have presented numerous Webservices composition methods Web service compositionmethods can be divided into workflow-based service compo-sitions and artificial intelligence-based (AI) service composi-tions in accordance with their technical and theoretical basesFor the workflow method a composite service is similar to aworkflow that contains a set ofWeb services together with thecontrol and data flow among the services (eg [33 34 4647]) For the AI method the Web service composition canbe regarded as an automatic method of finding the solutionto a planning problem given an initial state and the targetstate seek a path to achieve the service portfolio from theinitial state to the target state in a collection of services(eg [48ndash50]) The automatic and intelligent AI-based Webservice composition method develops the trend and the finalpurpose however the workflow method is more mature inindustry In this paper aWeb service composition frameworkbased on both of these methods is proposed

Chen et al [33 34] introduced a geospatial processingworkflow (GPW) method to integrate geo-related Web ser-vices for a complex task this method has some notableadvantages over other methods interoperability flexibilityand reusability The GPW method provides a general frame-work called the abstract GPW which defines the conceptionprocess of a task and instantiates a concrete GPW in aspecific application Abstract GPW consists of three phasesknowledge information and data [33]The knowledge phasedefines geospatial models and processes The informationphase integrates geospatial processes into a geospatial servicechain The data phase executes a geospatial service chainto generate data This paper focuses on the GPW methodfor GIS and RS Web service composition A framework forintegrating GIS and RS models based on Web service usingthe GPWmethod is proposed as depicted in Figure 4

Analogous with the roles in SOA there are three modelroles in the framework shown in Figure 4 themodel providerthe model broker and the model consumer

(i) Themodel provider publishesmodels toWeb services(described in Section 21) and registers them with themodel broker

(ii) Themodel consumer is a user who finds a solution fora task from the service broker

(iii) The model broker is a model metainformation repos-itory and a task solver

Core parts in the model broker are the geospatial knowl-edge the service repository and the geospatial processingengine

(i) Geospatial knowledge provides geospatial knowl-edge for intelligently processing model integration itincludes a semantics library and a geospatial processchain knowledge library

(ii) Service repository is a center that accepts the registra-tion of models in aWeb service format with semanticannotation

(iii) Geospatial processing engine is responsible for han-dling the task from the model consumer

The purpose of this framework is not only to enable themodel owner and the model consumer to perform little work


Knowledge phase

Geospatial knowledgeGeospatial process

chain knowledge library

Geospatialsemantics library

Atmosphericcorrection

Radiometriccorrection

Geometriccorrection

Chlorophyll ainversion

Task

Geospatial models and processes

Process chainknowledge

example

ModelsWeb services

(1) Searchknowledge library

to find suitableprocess chain to

the task

(2) Find suitableWeb services foreach process ofa process chain

(3) Form servicechains of the

process chains

Service broker

Service consumer Service provider

Model consumer Model provider

Find

Find Register

Register

Binding

Binding

Geospatialknowledge

Servicerepository

Geospatialprocessing engine

Model broker

Service chain

Information phase Data phase

Geospatial service chain Geospatial data

Result

Execution engine

Service chainoptimizationand selection

Data inconsistencyprocessing

(4) Postprocessservice chain

Service chaincomposition

Figure 4 Integrate GIS and RS models based on Web service under the GPW framework

but also to enjoy the model sharing and integration leavingthe sharing and integration challenge to the model brokerTo achieve this some strategies are incorporated in the threecore parts of the model broker based on the three phases ofthe GPW illustrated in Figure 4

221 Knowledge Phase Isolated geospatial models are notdesigned to be combined together Geospatial knowledgeincluding a geospatial process chain knowledge library and

a geospatial semantic library helps to combine models and isprepared by the model broker A model is regarded here asa process A process chain (CP) is an ordered logical modelcombination in semantics CP is formalized as a tuple as CP =ID 119873 119877MD where ID is the unique identification of theCP119873 is its name119877 is the relationship vectors for themodels119877 = 119872

11198722 119872

119894presents a model 119872 (mentioned in

the beginning of Section 2) with subscript 119894 (a finite number)and MD indicates the metadata of the CP For examplethe chlorophyll-a inversion process chain in Figure 4 is CP


Process of Web service integration

GIS service

RS service

Compositionservice

Webserviceparsingbased

onWSDL

Webservice

integrationbased oninterfaces

Webservice

interfaces(models

inputs andoutputs)

Figure 5 Processes of Web service integration in the model broker

the 119877 of the CP is 119877 = 1198721119872211987231198724 where 119872

1

11987221198723 and119872

4denote the atmospheric correction model

the radiometric correction model the geometric correctionmodel and the chlorophyll-a inversion model respectivelyTo effectively search a process chain the semantics librarycontains ontologies which formally represents knowledge asa set of concepts within a domain using shared vocabulary todenote the types properties and interrelationships of thoseconcepts [51] A semantics library provides the capabilitiesof eliminating the inconsistencies among process names andmodels names For an application task the responsibilities ofknowledge phase are to find suitable process chains andWebservices in three steps step (1) search the knowledge libraryto find a suitable process chain for the task step (2) findsuitable Web services for each process in the process chainand step (3) form service chains of the process chains In step(1) the semantic terms to describe a task are found or set bymodel consumer according to the semantics library providedby themodel brokerThere is a graphical user interface (GUI)developed by the model broker for the model consumer touse when submitting a task The model broker chooses atask from the ones that have been already listed in the GUIor submits a core term describing the task to the GUI andthen finds the task from the returns of the GUI The processchain library shares the same semantics libraryTherefore thework of step (1) is semantics matching Step (2) collects all theassociated models or Web services of each process and thenforms Web service chains of the process chains this step isautomatically completed by the model broker

222 Information Phase The original Web service chainsderived from the knowledge phase are mainly semanticchains The transition from the semantic chains to an exe-cutable workflow chain requires Web services compositiondata inconsistency processing and service chain optimiza-tion and selection as shown in Figure 4

Web Services Composition The processes of Web serviceintegration are shown in Figure 5 A Web service parsingengine is designed by the model broker to parse RS andGIS Web services with their WSDLs Next the Web serviceinterface extracts the appropriate models and composes themodels with theWSDL interfaces Figure 6 shows three basiccomposition relationships between two interfaces interfaceIName1 and interface IName2 with the pseudo-Unified

Modeling Language (UML) diagram If the output of anoperation in an interface is part or all of the input of theotherrsquos interface the two interfaces are associated as shownin Figure 6(a) Their composite is sequential If the outputsof two interfaces are the inputs of another interface the twointerfaces work together to form a collaborative relationshipas shown in Figure 6(b) In contrast with their concurrentappearance the chosen relationship chooses one interface fora further composite as shown in Figure 6(c) Based on thesebasic composite relationships many Web services integratetogether for each task

Data Inconsistency Processing The data flow of the Webservice composition is a chain of the inputs and the outputsof the models The core model executions are black boxesbut the associatedmodels do not need to have consistent dataformats For example in Figure 6(a) the outputs of IName1are the logical inputs of IName2 but the formats may not bephysically consistent because one is in GoTIFF data formatand the other is in IMG data format The data inconsistencycan be caused by themodelerrsquos preference for setting the inputand output parameters Thus the compositing Web serviceis not only determining the workflow of the Web servicesbut is also processing the data inconsistencies Vector andraster data are the two basic and most commonly used datatypes in GIS and RS Data inconsistencies of GIS and RS canbe external and internal as shown in Table 1 To overcomethis problem transformation functions are usedWeb servicesas shown in Figure 7 Data type transformation coordi-nate system transformation data format transformationand resolution transformation are necessary transformationfunctions These four functions are basic functions in GISand RS Widely used professional software includes thesefunctions For example the ESRI ArcGIS ArcToolbox pro-vides hundreds of geospatial-related analysis and processingfunctions including these four transformation functionsThetransformation functions from existing professional software(eg ESRI ArcGIS ArcToolbox) have been published intothe Web services by the model broker with the methodmentioned in Section 21 and have been registered in themodel broker

To eliminate data inconsistencies automatically somerules are defined to register the inputs and the outputs ofthe functions of a Web service by the model broker If someparameters of the inputs and the outputs of a function aredata the parametersmust be described using their properties


ldquoInterfacerdquo ldquoInterfacerdquo


IName1 IName2

IName1IName2OpName(Inputs) Outputs OpName(Inputs) Outputs

OpName(Inputs) Outputs OpName(Inputs) Outputs

(a) Associated relationship

IName1

IName3

IName2

1lowast1

1lowast1

OpName(Inputs) Outputs


ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo

(b) Collaborated relationship

IName1OpName(Inputs) Outputs


IName3Choose

IName2

ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo

(c) Chosen relationship

Figure 6 Basic composition relationships between two interfaces

Data inconsistency transformation services


Data typetransformation

Web service

Coordinatesystem

transformationWeb service

Resolutiontransformation

Web service

Data formattransformation

Web service

IName1


IName2

SOAP

Web service

SOAP

Web service

SOAP

Web service

SOAP

Web service


Figure 7 Data inconsistency transformation services

Table 1 Data inconsistency of two data types

Data 1 Data 2 Presentation of data inconsistencyVector data Raster data Data typeRaster data Vector data Data typeVector data Vector data Coordinate system

Raster data Raster dataCoordinate system

ResolutionData format

data type (DT) satellitesensor type (ST) coordinate systems(CS) resolution (RE) and data format (DF) The value of

DT is ldquovectorrdquo or ldquorasterrdquo The ST is the satellitesensor thatobtained the data such as ldquoMODISrdquo [52] or ldquoTerraSAR-Xrdquo[53] The representation of CS meets the requirement of thePROJ4 [54] RE is numericDF is a commondata format suchas ldquoGeoTiffrdquo Knowing the five aspects of data inconsistenciesinconsistent data can be transformed to consistent formatsFor eachmodel themodel provider describes the five aspectsaccording to the rules defined by model broker

Service Chain Optimization and Selection To improve effi-ciency the integration processes of the Web services need tobe optimized and selected Assume there is a task TASK

119899

that has a series of Web service composition schemes


SCHEME1 SCHEME

119899 The cost of the 119894th (1 le 119894 le

119899) scheme is 119879119894 The objective is to determine the scheme

with the minimal cost Min(119879119894) Zeng et al [55] proposed

five generic quality criteria for elementary services execu-tion price execution duration reputation reliability andavailability they also selected a global optimal executionscheme According to the features of this framework theglobal optimal execution time is a primary considerationThe final service chain schemes are the outputs of the modelbroker

223 Data Phase This is a result phase In this phasetwo types of resultsmdashdata results and workflow schemesresultsmdashare provided by the model broker For the formera geospatial service chain is executed by execution engine inthe model broker to obtain the desired data products Forthe latter the model broker outputs include the workflowschemes described with the service chains the associatedWeb services and the workflow descriptions The differencebetween the two is where the workflow is executed theformer is executed at the model broker and the latter isexecuted at the model consumer The Business Process Exe-cution Language (BPEL) [56] is an advancing open standardsfor the information society standard executable languagefor specifying actions within business processes with Webservices The workflow described using BPEL is flexibleand reusable [33 34] Therefore an executable workflow isdescribed using BPEL

3 Example Case and Result

The model sharing and integration method based on Webservice that is proposed in this paper is applied for the waterenvironment monitoring in the Pearl River Delta (PRD)region which is introduced in Section 31 Then the resultsof this application including the publication of the model asaWeb service and the integration of themodelsrsquoWeb servicesare executed evaluated and discussed in Sections 32 and 33

31 Pearl River Delta Water Environment Monitoring ThePRD located between latitudes 21∘401015840N and 23∘N andbetween longitudes 112∘E and 113∘201015840E is the low-lying areasurrounding the Pearl River estuary in China where thePearl River flows into the South China Sea The PRD is aregion in China experiencing one of the fastest economygrowth rates from Chinarsquos reformation and opening in 1979Many largemetropolises such as Guangdong and the specialadministrative regions of Shenzhen Macau and Hong Kongare nearby With rapid economic development and urbaniza-tion water environment problems such as water pollutionand water safety are becoming serious concerns in the PRD[57ndash60] To protect the water environment for better livingconditions and sustainable development governments in thePRD area initiated several programs including developinga water environment monitoring system Seven researchinstitutesuniversities with scholars from a range of scientificdomains such as hydrology ecology RS and GIS are collab-orating together to develop a water environment monitoringsystem for the PRD

Table 2 Models and their runtime environments

Model name Platform Language Execution methodAtmosphericcorrection model

MicrosoftWindows ENVI IDL ENVI script (pro)

Chlorophyll-ainversion model

MicrosoftWindows PCI IDL PCI script (eas)

Projectiontransformationmodel

MicrosoftWindows C EXE

Because it is a collaborative effort an initial problemis that the models are scattered in different areas withdistributed systems in the PRD In addition some modelsare not completely open This is because some models areconsidered to be core secrets in their institutes and othermodels are authorization-required and classified Thereforesome models are not easy to obtain and model owners preferto provide the ldquofinal productrdquo derived from themodels ratherthan the models themselves In addition there are differenttypes of models (RS models and GIS models) differentrunning platforms (Linux and Window) and different pro-gramming languagesscripts ENVI IDL [61] and PCI EASI[62] which make them difficult to integrate

To overcome the existing problems in the PRD waterenvironment monitoring system the system functions areperformed by sharing and integrating the RS and GIS modelsbased on a Web service as shown in Figure 8(a) Finallythe water environment monitoring system is developedincluding the functions shown in Figure 8(a) the systemframework shown in Figures 8(b)-8(c) and the portal shownin Figure 8(d)

32 Publishing a Model as a Web Service The purpose ofthis experiment is to demonstrate the results of the PRDwater environment system using the Web service modelsharing method mentioned in Section 21 The experimentwas performed by codevelopers of the PRD system fromseveral institutesuniversities The codevelopers publishedtheir ownmodels using theWeb service publishing platformA subset of the models with their runtime environments arelisted as examples in Table 2The execution methods listed inTable 2 indicate the programming entry of an encapsulatedmodel From the table it is evident that models with differentdevelopment languages and different execution methods arepublished into the Web services

33 Integrating Models as Web Services The aim of thisexperiment is to show the modelsrsquo integration using the Webservices Building geospatial knowledge managing a task andselecting a reasonable result are each explained

Building Geospatial Knowledge Building geospatial knowl-edge consists of creating the geospatial process chain knowl-edge library and the geospatial semantics library as shownin Figure 4 The process chains are collected from existingprofessional software and textbooks For example the GISprofessional software ArcGIS toolbox has already provided


Application layer

Logical layer

Data andservice layer

Representationlayer

Representation

Link

Associate

Function modules

Data base Map service

Portal

Users

Project title

Map tool bar

Map

Function menus

Function panel

(d) System portal middot middot middot

Representation

Link

ArcGIS

Oracle databaseFundamental

geographic dataObserved data

Remote sensing data

ASPNET Silverlight

DistributedWeb servicesPCI OSG

Developed tool

Web service

ArcGIS server

ArcTools

Composited Web services

Internet

server

(c) System framework implementation

Land and water separation

Extraction of lakersquos width and length

Measurement of water quantity in lake

Extraction of riverrsquos width and length

River flows

Cloud detection and image mosaic


Radiometric correctionAtmospheric correction

Measurement of saltwater intrusion

Fusion between image data and measured data

Evaluations of water qualityrsquos classification

Statistical report

Presentation of video and image

Geometric correction

Batch processing of remotesensing data

(1) Layersrsquo representation and operation module

Measurement of chlorophyll aMeasurement of suspended sediment

Organic pollutantsOil pollution

Pollution source detectionWater quality classification

Image result fusion

(2) Image data management module(3) Real data management module

(5) Remote sensing monitoring of water environment in land

(4) Image data pre-processing module

Image data fusion

Fusion between image result and real data



Radiometric correction

Atmospheric correction

Geometric correctionBatch processing of remote sensing

data

(b) Abstract system framework


Measurement of yellow substanceSea surface temperature

Transparency

(6) Remote sensing monitoring of water environment in sea

(7) Resultsrsquo representation moduleSearch of water quality and quantity

(a) System functionsSimulation analysis of 2D and 3D scene

Rem

ote s

ensin

g m

onito

ring

syste

m o

f wat

er en

viro

nmen

t

Mea

sure

men

t of

wat

er q

ualit

yM

easu

rem

ent o

fw

ater

qua

lity

Mea

sure

men

t of

wat

er q

ualit

y

Imag

e dat

a bef

ore-

proc

essin

gin

land

mod

ule

Imag

e dat

a pre

-pro

cess

ing

in se

a mod

ule

Mul

ti-im

age

data

fusio

n

Figure 8 The system functions and system portal


Raw data Web service

Database

System

txtALTOVA

HJ1A-CCD1-456-90-20101108-L20000855826JPG

HJ1A-CCD1-456-90-20101108-L20000855826XML

HJ1A-CCD1-456-90-20101108-L20000855826-1TIF

HJ1A-CCD1-456-90-20101108-L20000855826-2TIF

HJ1A-CCD1-456-90-20101108-L20000855826-3TIF

HJ1A-CCD1-456-90-20101108-L20000855826-4TIF

HJ1A-CCD1-456-90-20101108-L20000855826-SatAn

gletxt

HJ1A-CCD1-456-90-20101108-L20000855826-THU

MBJPG

Figure 9 The result of scheme 1

19 top-level tools with hundreds of functions RS professionalsoftware ENVI ERDAS and PCI also provide hundreds offunctions Then semantic annotations are made for eachprocess chain Currently the semantic annotation of eachprocess in a chain is based on the name and its functionalrelationship

Managing a Task The main work of managing a task is tomatch a taskwith a process and itsWeb servicesThe semanticmatch of a task with a process chain is based on the semanticterm For example if a task is to calculate chlorophyll athen the task with the semantic term ldquochlorophyll ardquo andthe process chain with semantic term ldquochlorophyll ardquo arematched Because the semantics and the Web services of aprocess are registeredwith fixed standards at the beginning bythe model providers the suitable Web services for a processchain will be found The data inconsistency processing

includes transforming data to another data format (DT STCS RE and DF)

Selecting a Reasonable Service Chain The result of the modelbroker is a data result or a set of model integration schemesIn the PRD system the model integration schemes arechosen The model broker shows the reasonable results asa list for example a scheme for separating land and wateris introduced Separating land and water is a critical stepfor extracting water area and the water environment Theraw data are processed radiometric correction atmosphericcorrection and image segmentation to separate land andwater Two resulting schemes are return from the modelbroker and two examples are tested The data in example 1is a 259MB TM image the data in example 2 is an 826MBHJ image The time costs of the two examples are providedin Table 3 As shown in Table 3 the costs of schemes 1


Table 3 The cost comparisons of two Web services methods

Scheme name Cost (second)Example 1 Example 2

Scheme 1 534840 1012558Scheme 2 648795 227876Ratio 8244 4443

and 2 are 8244 and 4443 respectively Therefore theefficiency of scheme 1 is higher and scheme 1 is more highlyrecommended than scheme 2The result of scheme 1 is shownin Figure 9

4 Discussion and Conclusion

Sharing and integrating RS and GISmodels are important fortheir applicationThis paper studies the framework of sharingand integrating RS and GIS models using a Web service Inthe framework a black box method with a visual interfaceis proposed for rapid Web service publishing This methodwill facilitate the publication of a model to the Web servicein addition it will assist researchers in concentrating theirefforts on model development rather than on programmingIn addition integrating workflow and using semantics sup-ported method is an effective way to integrate models usingWeb services The framework is applied in the developmentof the PRD water environment monitoring system in whichRS and GIS models are integrated

The advanced features of this framework are facilitatingthe model providerrsquos and model consumerrsquos work of modelsharing and integration and integrating the workflow-basedservice composition method and the AI-based service com-position method For the former the framework providesthe model provider and the model consumer methods forrapidly publishing and integrating their models A modelprovider can publish a model using a visual interface anda model consumer chooses the semantic terms of the taskThis frees the model providers and the model consumersfrom tediouswork and improves their work efficiency For thelatter the framework integrates the workflow-based servicecomposition method and the AI-based service compositionmethod making model integration intelligent automaticand industrialized

Further studies will focus on the following aspects

(i) enhancing and enriching the geospatial knowledgethe geospatial knowledge in the model broker isderived from professional software and textbooksThe process chains and semantic information arelimited requiring enhancement and enrichment tomeet the requirement of broader applications

(ii) improving the automatic processing ability the auto-matic processing ability depends on the geospatialknowledge and the Web services composition Cur-rently the Web services composition is associatedwith the process chains More powerful and efficientassociation between a process chain and its Webservices require additional work

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

This work has been supported by the National Key Tech-nology RampD Program of China (no 2012BAH32B03) HongKong ITF Program (no GHP00211GD) Hong Kong ITFProgram (no ITS04212FP) and the National Natural Sci-ence Foundation of China (NSFC) Program (no 41301441)

References

[1] Q Weng ldquoModeling urban growth effects on surface runoffwith the integration of remote sensing and GISrdquo EnvironmentalManagement vol 28 no 6 pp 737ndash748 2001

[2] C Granell L Dıaz and M Gould ldquoService-oriented applica-tions for environmental models reusable geospatial servicesrdquoEnvironmental Modelling and Software vol 25 no 2 pp 182ndash198 2010

[3] J C Hinton ldquoGIS and remote sensing integration for envi-ronmental applicationsrdquo International Journal of GeographicalInformation Systems vol 10 no 7 pp 877ndash890 1996

[4] C Min L GuonianW Yongning T Hong and F Guo ldquoStudy-ing on distributed sharing of geographical analysis modelrdquo inProceedings of theWRIWorld Congress on Computer Science andInformation Engineering (CSIE rsquo09) pp 346ndash349 April 2009

[5] Y Wen M Chen G Lv H Lin L He and S Yue ldquoPrototypingan open environment for sharing geographical analysis modelson cloud computing platformrdquo International Journal of DigitalEarth vol 6 no 4 pp 356ndash382 2013

[6] MChen YH Sheng YNWenH Tao and FGuo ldquoSemanticsguided geographic conceptual modeling environment based oniconsrdquo Geographical Research no 3 pp 705ndash715 282009

[7] H LinMChenG Lv et al ldquoVirtualGeographic Environments(VGEs) a new generation of geographic analysis toolrdquo Earth-Science Reviews vol 126 pp 74ndash84 2013

[8] H Lin M Chen and G Lv ldquoVirtual Geographic Environment-a workspace for computer-aided geographic experimentsrdquoAnnals of the Association of American Geographers vol 103 no3 pp 465ndash482 2013

[9] A Brooke D Kendrick and EMeerausGAMS A Users GuideThe Scientific Press Redwood Calif USA 1988

[10] R Fourer D M Gay and B W Kernighan AMPL A ModelingLanguage for Mathematical Programming The Scientific PressRedwood Calif USA 1993

[11] S Katz L J Risman and M Rodeh ldquoA system for constructinglinear programming modelsrdquo IBM Systems Journal vol 19 no4 pp 505ndash520 1980

[12] H K Bhargava and S O Kimbrough ldquoEmbedded languages formodelmanagementrdquoDecision Support Systems vol 10 no 3 pp277ndash299 1993

[13] R W Blanning ldquoA relational framework for join implementa-tion in model management systemsrdquo Decision Support Systemsvol 1 no 1 pp 69ndash85 1985

[14] T-P Liang ldquoIntegratingmodel management with datamanage-ment in decision support systemsrdquo Decision Support Systemsvol 1 no 3 pp 221ndash232 1985


[15] H Bunke ldquoAttributed programmed graph-grammars and theirapplications to schematic diagram interpretationrdquo IEEE Trans-actions on Pattern Analysis and Machine Intelligence vol 4 no6 pp 574ndash582 1982

[16] C V Jones ldquoIntroduction to Graph-Based Modeling SystemsPart II Graph-grammars and the implementationrdquo ORSAJournal of Computing vol 3 no 3 pp 180ndash206 1991

[17] M Chen Y Sheng Y Wen H Tao and F Guo ldquoSemanticguided geographic conceptual modeling environment based oniconsrdquo Geographical Research vol 28 no 3 pp 705ndash715 2009

[18] M Chen H Tao H Lin and Y Wen ldquoA visualization methodfor geographic conceptual modellingrdquoAnnals of GIS vol 17 no1 pp 15ndash29 2011

[19] A M Geoffrion ldquoAn introduction to structured modelingrdquoManagement Science vol 33 no 5 pp 547ndash588 1987

[20] A M Geoffrion ldquoThe SML language for structured modelinglevels 1 and 2rdquoOperations Research vol 40 no 1 pp 38ndash57 1992

[21] O El-Gayar and K Tandekar ldquoAn XML-based schema defini-tion for model sharing and reuse in a distributed environmentrdquoDecision Support Systems vol 43 no 3 pp 791ndash808 2007

[22] CORBA Component Model Specification version 40 ObjectManagement Group httpwwwomgorgspecCCM40

[23] Distributed Component Object Model (DCOM) RemoteProtocol Microsoft httpdownloadmicrosoftcomdown-load95E95EF66AF-9026-4BB0-A41D-A4F81802D92C[MS-DCOM]pdf

[24] T B Downing Java RMI Remote Method Invocation IDGBooks Worldwide Foster City Calif USA 1998

[25] G Bucci F Ciancetta E Fiorucci D Gallo and C Landi ldquoAlow cost embedded web services for measurements on powersystemrdquo in Proceedings of the IEEE International Conference onVirtual Environments Human-Computer Interfaces and Mea-surement Systems (VECIMS rsquo05) pp 1ndash6 2005

[26] MDA Guide Version 101 Object Management Grouphttpwwwomgorgcgi-bindocomg03-06-01pdf

[27] Model-Driven Architecture Vision Standards And Emerg-ing Technologies Object Management Group httpwwwomgorgmdamda filesModel-Driven Architecturepdf

[28] J L Goodall B F Robinson and A M Castronova ldquoModelingwater resource systems using a service-oriented computingparadigmrdquo Environmental Modelling and Software vol 26 no5 pp 573ndash582 2011

[29] T Maxwell and R Costanza ldquoAn open geographic modelingenvironmentrdquo Simulation vol 68 no 3 pp 175ndash185 1997

[30] T Maxwell and R Costanza ldquoA language for modular spatio-temporal simulationrdquo Ecological Modelling vol 103 no 2-3 pp105ndash113 1997

[31] M Reed S M Cuddy and A E Rizzoli ldquoA frameworkfor modelling multiple resource management issuesmdashan openmodelling approachrdquo Environmental Modelling and Softwarevol 14 no 6 pp 503ndash509 1999

[32] M-H Tsou and B P Buttenfield ldquoA dynamic architecture fordistributing geographic information servicesrdquo Transactions inGIS vol 6 no 4 pp 355ndash381 2002

[33] N Chen L Di G Yu and J Gong ldquoGeo-processing workflowdriven wildfire hot pixel detection under sensor web environ-mentrdquo Computers and Geosciences vol 36 no 3 pp 362ndash3722010

[34] N Chen L Di G Yu and J Gong ldquoAutomatic on-demand datafeed service for autochem based on reusable geo-processing

workflowrdquo IEEE Journal of Selected Topics in Applied EarthObservations and Remote Sensing vol 3 no 4 pp 418ndash4262010

[35] A Grimshaw M Morgan D Merrill et al ldquoAn open gridservices architecture primerrdquo Computer vol 42 no 2 pp 27ndash34 2009

[36] L Dıaz C Granell M Gould and V Olaya ldquoAn open servicenetwork for geospatial data processingrdquo in An Open ServiceNetwork for Geospatial Data Processing Free and Open SourceSoftware for Geospatial (FOSS4G) Conference pp 410ndash4202008

[37] S Nativi P Mazzetti and G N Geller ldquoEnvironmental modelaccess and interoperability the GEO Model Web initiativerdquoEnvironmental Modelling and Software vol 39 pp 214ndash2282013

[38] D Roman S Schade A J Berre N R Bodsberg and JLanglois ldquoModel as a service (MaaS)rdquo inAGILEWorkshop GridTechnologies for Geospatial Applications 2009

[39] G N Geller and F Melton ldquoLooking forward Applying anecological model web to assess impacts of climate changerdquoBiodiversity vol 9 no 3-4 pp 79ndash83 2008

[40] G N Geller and W Turner ldquoThe model Web a conceptfor ecological forecastingrdquo in IEEE International Geoscience ampRemote Sensing Symposium (IGARSS rsquo07) pp 2469ndash2472 June2007

[41] M F Goodchild GIS Spatial Analysis and Modeling ESRIPress Redlands Calif USA 2005

[42] B Srivastava and J Koehler ldquoWeb service composition-currentsolutions and open problemsrdquo inWorkshop on Planning forWebServices (ICAPS rsquo03) pp 28ndash35 2003

[43] P Yue L Di W Yang G Yu and P Zhao ldquoSemantics-based automatic composition of geospatialWeb service chainsrdquoComputers and Geosciences vol 33 no 5 pp 649ndash665 2007

[44] M Rouached W Fdhila and C Godart ldquoWeb services com-positionsmodelling and choreographies analysisrdquo InternationalJournal of Web Services Research vol 7 no 2 pp 87ndash110 2010

[45] B BeizerBlack-Box Testing Techniques For Functional Testing ofSoftware and Systems JohnWiley amp Sons New York NY USA1995

[46] F Casati and M-C Shan ldquoEvent-based interaction manage-ment for composite e-services in eflowrdquo Information SystemsFrontiers vol 4 no 1 pp 19ndash31 2002

[47] F Casati S Ilnicki L J Jin V Krishnamoorthy and M CShan ldquoAdaptive and dynamic service composition in eFlowrdquo inAdvanced Information Systems Engineering vol 1789 of LectureNotes in Computer Science pp 13ndash31 2000

[48] S McIlraith and T C Son ldquoAdapting golog for composition ofsemantic web servicesrdquo KR vol 2 pp 482ndash493 2002

[49] J Peer ldquoA PDDL based tool for automatic web service composi-tionrdquo in Principles and Practice of Semantic Web Reasoning vol3208 ofLectureNotes inComputer Science pp 149ndash163 SpringerBerlin Germany 2004

[50] E Sirin B Parsia D Wu J Handler and D Nau ldquoHTNplanning for web service composition using SHOP

2rdquo Web

Semantics vol 1 no 4 pp 377ndash396 2004[51] T R Gruber ldquoA translation approach to portable ontology

specificationsrdquoKnowledge Acquisition vol 5 no 2 pp 199ndash2201993

[52] L A Remer Y J Kaufman D Tanre et al ldquoTheMODIS aerosolalgorithm products and validationrdquo Journal of the AtmosphericSciences vol 62 no 4 pp 947ndash973 2005


[53] R Werningh ldquoTerraSAR-X missionrdquo in Remote Sensing Inter-national Society for Optics and Photonics pp 9ndash16 2004

[54] Cartographic Projection Procedures for the UNIXEnvironment-A Usersquos Manual United States Departmentof the Interior Geological Survey ftpftpremotesensingorgprojOF90-284pdf

[55] L Zeng B BenatallahMDumas J Kalagnanam andQ ShengldquoQuality driven web services compositionrdquo in Proceedings of the12th International Conference on World Wide Web pp 411ndash4212003

[56] Web Services Business Process Execution LanguageVersion 20OASIS httpswwwoasis-openorgcommitteesdownloadphp21575wsbpel-specification public review draft 2 diffpdf

[57] X Fan B Cui H Zhao Z Zhang and H Zhang ldquoAssessmentof river water quality in Pearl River Delta using multivariatestatistical techniquesrdquo Procedia Environmental Sciences vol 2pp 1220ndash1234 2010

[58] Y Zhang Y Wang Y Wang and H Xi ldquoInvestigating theimpacts of landuse-landcover (LULC) change in the pearl riverdelta region onwater quality in the pearl river estuary andHongKongrsquos coastrdquo Remote Sensing vol 1 no 4 pp 1055ndash1064 2009

[59] N Zhou B Westrich S Jiang and Y Wang ldquoA couplingsimulation based on a hydrodynamics and water quality modelof the Pearl River Delta Chinardquo Journal of Hydrology vol 396no 3-4 pp 267ndash276 2011

[60] T Ouyang Z Zhu and Y Kuang ldquoAssessing impact of urban-ization on river water quality in the Pearl River Delta EconomicZone Chinardquo Environmental Monitoring and Assessment vol120 no 1ndash3 pp 313ndash325 2006

[61] M J Canty Image Analysis Classification and Change Detectionin Remote Sensing With Algorithms For ENVIIDL CRC PressBoca Raton Fla USA 2007

[62] Geomatica EASI User Guide(version 101) PCI Geomaticshttpwwwgisunbccahelpsoftwarepcieasipdf

Submit your manuscripts athttpwwwhindawicom

Computer Games Technology

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Distributed Sensor Networks


Advances in

FuzzySystems

Hindawi Publishing Corporationhttpwwwhindawicom

Volume 2014


ReconfigurableComputing

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014


Applied Computational Intelligence and Soft Computing

thinspAdvancesthinspinthinsp

Artificial Intelligence

HindawithinspPublishingthinspCorporationhttpwwwhindawicom Volumethinsp2014

Advances inSoftware EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Electrical and Computer Engineering

Journal of

Journal of

Computer Networks and Communications


Hindawi Publishing Corporation

httpwwwhindawicom Volume 2014

Advances in

Multimedia


Biomedical Imaging


ArtificialNeural Systems

Advances in


RoboticsJournal of



Computational Intelligence and Neuroscience

Industrial EngineeringJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Human-ComputerInteraction

Advances in

Computer EngineeringAdvances in



the Common Object Request Broker Architecture (CORBA)[22] the Distributed Component Object Model (DCOM)[23] and the Java Remote Method Invocation (Java RMI)[24] the direct use of these three technologies by pro-grammers is becoming less common for their shortcomingstightly coupled protocols are effective for building a specificapplication but are not sufficiently flexible and protocols areconstrained by their vendor implementations platforms andlanguages [25]

In recent years standard technologies specifications andframeworks have been developed for model sharing andintegration examples include theModel Driven Architecture(MDA) and the open modeling interface (OpenMI) as wellas several modeling frameworks MDA is a software designapproach developed by the Object Management Group toprovide an open and vendor-neutral approach for softwaresystems using specifications such as The Unified ModelingLanguage the Meta Object Facility the XML MetadataInterchange and the Common Warehouse Meta-model thatbuilt coherent schemes for authoring publishing and man-aging models with the platform independent model PIMand the platform-specific model PSM [26 27] The MDA ispromising but the PIM and the PSM do not include standardspecification mapping which complicates application of theMDA The open modeling interface (OpenMI) is a softwarecomponent interface that allows time-dependent modelsto share and exchange information at runtime makingmodel integration feasible at the operational level OpenMIemphasizes interoperability between otherwise independentmodels [28] Dozens of modeling and reuse frameworkssuch as the spatial modelling environment (SME) [29 30]and the interactive component modelling system (ICMS)[31] have been developed to handle model sharing andintegration problems but most of them are highly field-related platform-related and internet-inaccessible whichlimits their application [5]

With the development of information technology (IT)Web service and service-oriented architecture (SOA) aremore commonly used in model sharing and model integra-tion [2 32ndash34] Some advantages include sharing models asworldwide services applied to various popular applications[35] and integrating data and models from anywhere in theworld [36] In addition with the advent of the advancednotion of ldquoModel as a Servicerdquo [37ndash40] the Web servicemethod creates new opportunities for sharing and integratingmodels including GIS and RS models

The objective of this paper is to design a Web service-based method to share and integrate RS and GIS modelsThemain idea of model sharing and integration based on Webservice is to release various heterogeneousRS andGISmodelsas Web services for model sharing and then to integrate thehomogeneousWeb services for model integration Followingthis idea this paper publishes the methods model and themodel integration as Web services and applies this to thewater environmentmonitoring in the Pearl RiverDelta regionin China

The organization of this paper is as follows Section 2addresses the main idea of web service-based model shar-ing and integration and the proposed method Section 3

describes the implementation of the proposed method forthe PRD water environment monitoring system FinallySection 4 concludes this paper and outlines future work

2 Methodologies

A model is a representation of one or more processes thatare believed to occur in the real word [41] From a userrsquosperspective a model can be formalized as a tuple as 119872 =ID 119873 119868 119874 IMPLMD where119872 is the model ID presentsthe unique identification of themodel119873 is its name and 119868119874IMPL andMD are the input the output the implementation(executable programs) and the metadata of the modelrespectively

Web service offers a way to share and interoperatemodels[40] The method is based on Simple Object Access Protocol(SOAP) and Web Services Description Language (WSDL)WSDL provides a function-centric description of the Webservices including their inputs outputs and exception hand-ing [42] which map to the structure of the formalized modeltuple Because the method is standardized and loosely cou-pled [40] heterogeneous RS andGISmodels become ldquohomo-geneousrdquo Web services after they are published Integratingheterogeneous RS andGISmodels is equivalent to integratingthe ldquohomogeneousrdquo Web services because the models arepublished as Web services Web service composition is amethod of integrating Web services [33 34 43] Web servicecomposition aggregates multiple services into a single newservice for a certain functionality that a single primary servicecannot provide [44] If a model that runs on a desktop orlocal area network environment is called the local model anda model that is published with Web service is called a Webmodel the aim of sharing and integrating the GIS and RSmodels is to publish the local model to a Web model andthen combine theWebmodelsThe framework of sharing andintegratingGIS andRSmodels based onWeb service is shownin Figure 1

21 Model Sharing with Web Service Model sharing is thepublishing of local models as Web models using a Webservice Developing a Web service is time intensive andrequires specific professional skills In addition many GISand RS model experts aim to concentrate their efforts ondeveloping more powerful models rather than on improvingtheir programming skills Considering these factors a blackbox approach combined with a visual method is introducedto facilitate the publishing of local models to Web services Ablack box [45] is a method that can be used without knowinghow its inner algorithm works the user only needs to knowthe input and output characteristics The exposed entry ofthe black box is a visual user interface that is illustrated inFigure 2 A visual method for manipulating a model is aneffective way to reduce development time and training [46]The visual interface hides the cumbersome implementationof the Web service

For this method ldquoinsiderdquo the black box is criticalgenerating Web services from GIS and RS models withtheir names inputs outputs and executable programs The


GIS models

RS models

Localserver




Heterogeneousmodels

Homogeneousservice

Procedure

Localserver

Localserver

SOAP

Web service

SOAP

Web service

SOAP

Web service


RS

Black box

Web service

GIS model

model








Model

executioncode

ModelInput

Output

Web service

WSDL


Execution code



















Knowledge phase






Geometriccorrection


Task



example

ModelsWeb services



the task



process chains

Service broker



Find

Find Register

Register

Binding

Binding

Geospatialknowledge

Servicerepository


Model broker

Service chain



Result

Execution engine









11198722 119872





GIS service

RS service

Compositionservice


onWSDL

Webservice


Webservice

interfaces(models

inputs andoutputs)



1

11987221198723 and119872











IName1 IName2




IName1

IName3

IName2

1lowast1

1lowast1



ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo




IName3Choose

IName2

ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo






Web service

Coordinatesystem



Web service


Web service

IName1


IName2

SOAP

Web service

SOAP

Web service

SOAP

Web service

SOAP

Web service










119899



SCHEME1 SCHEME






















Application layer

Logical layer


Representationlayer

Representation

Link

Associate

Function modules


Portal

Users

Project title

Map tool bar

Map

Function menus

Function panel


Representation

Link

ArcGIS



Remote sensing data

ASPNET Silverlight


Developed tool

Web service

ArcGIS server

ArcTools


Internet

server






River flows







Statistical report








Image result fusion




Image data fusion







data




Transparency




Rem

ote s

ensin

g m

onito

ring

syste

m o

f wat

er en

viro

nmen

t

Mea

sure

men

t of

wat

er q

ualit

yM

easu

rem

ent o

fw

ater

qua

lity

Mea

sure

men

t of

wat

er q

ualit

y

Imag

e dat

a bef

ore-

proc

essin

gin

land

mod

ule

Imag

e dat

a pre

-pro

cess

ing

in se

a mod

ule

Mul

ti-im

age

data

fusio

n




Database

System

txtALTOVA

HJ1A-CCD1-456-90-20101108-L20000855826JPG

HJ1A-CCD1-456-90-20101108-L20000855826XML

HJ1A-CCD1-456-90-20101108-L20000855826-1TIF

HJ1A-CCD1-456-90-20101108-L20000855826-2TIF

HJ1A-CCD1-456-90-20101108-L20000855826-3TIF

HJ1A-CCD1-456-90-20101108-L20000855826-4TIF

HJ1A-CCD1-456-90-20101108-L20000855826-SatAn

gletxt

HJ1A-CCD1-456-90-20101108-L20000855826-THU

MBJPG



















Acknowledgments


References





















































2rdquo Web






















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in




GIS models

RS models

Localserver




Heterogeneousmodels

Homogeneousservice

Procedure

Localserver

Localserver

SOAP

Web service

SOAP

Web service

SOAP

Web service


RS

Black box

Web service

GIS model

model








Model

executioncode

ModelInput

Output

Web service

WSDL


Execution code



















Knowledge phase






Geometriccorrection


Task



example

ModelsWeb services



the task



process chains

Service broker



Find

Find Register

Register

Binding

Binding

Geospatialknowledge

Servicerepository


Model broker

Service chain



Result

Execution engine









11198722 119872





GIS service

RS service

Compositionservice


onWSDL

Webservice


Webservice

interfaces(models

inputs andoutputs)



1

11987221198723 and119872











IName1 IName2




IName1

IName3

IName2

1lowast1

1lowast1



ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo




IName3Choose

IName2

ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo






Web service

Coordinatesystem



Web service


Web service

IName1


IName2

SOAP

Web service

SOAP

Web service

SOAP

Web service

SOAP

Web service










119899



SCHEME1 SCHEME






















Application layer

Logical layer


Representationlayer

Representation

Link

Associate

Function modules


Portal

Users

Project title

Map tool bar

Map

Function menus

Function panel


Representation

Link

ArcGIS



Remote sensing data

ASPNET Silverlight


Developed tool

Web service

ArcGIS server

ArcTools


Internet

server






River flows







Statistical report








Image result fusion




Image data fusion







data




Transparency




Rem

ote s

ensin

g m

onito

ring

syste

m o

f wat

er en

viro

nmen

t

Mea

sure

men

t of

wat

er q

ualit

yM

easu

rem

ent o

fw

ater

qua

lity

Mea

sure

men

t of

wat

er q

ualit

y

Imag

e dat

a bef

ore-

proc

essin

gin

land

mod

ule

Imag

e dat

a pre

-pro

cess

ing

in se

a mod

ule

Mul

ti-im

age

data

fusio

n




Database

System

txtALTOVA

HJ1A-CCD1-456-90-20101108-L20000855826JPG

HJ1A-CCD1-456-90-20101108-L20000855826XML

HJ1A-CCD1-456-90-20101108-L20000855826-1TIF

HJ1A-CCD1-456-90-20101108-L20000855826-2TIF

HJ1A-CCD1-456-90-20101108-L20000855826-3TIF

HJ1A-CCD1-456-90-20101108-L20000855826-4TIF

HJ1A-CCD1-456-90-20101108-L20000855826-SatAn

gletxt

HJ1A-CCD1-456-90-20101108-L20000855826-THU

MBJPG



















Acknowledgments


References





















































2rdquo Web






















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in




Model

executioncode

ModelInput

Output

Web service

WSDL


Execution code



















Knowledge phase






Geometriccorrection


Task



example

ModelsWeb services



the task



process chains

Service broker



Find

Find Register

Register

Binding

Binding

Geospatialknowledge

Servicerepository


Model broker

Service chain



Result

Execution engine









11198722 119872





GIS service

RS service

Compositionservice


onWSDL

Webservice


Webservice

interfaces(models

inputs andoutputs)



1

11987221198723 and119872











IName1 IName2




IName1

IName3

IName2

1lowast1

1lowast1



ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo




IName3Choose

IName2

ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo






Web service

Coordinatesystem



Web service


Web service

IName1


IName2

SOAP

Web service

SOAP

Web service

SOAP

Web service

SOAP

Web service










119899



SCHEME1 SCHEME






















Application layer

Logical layer


Representationlayer

Representation

Link

Associate

Function modules


Portal

Users

Project title

Map tool bar

Map

Function menus

Function panel


Representation

Link

ArcGIS



Remote sensing data

ASPNET Silverlight


Developed tool

Web service

ArcGIS server

ArcTools


Internet

server






River flows







Statistical report








Image result fusion




Image data fusion







data




Transparency




Rem

ote s

ensin

g m

onito

ring

syste

m o

f wat

er en

viro

nmen

t

Mea

sure

men

t of

wat

er q

ualit

yM

easu

rem

ent o

fw

ater

qua

lity

Mea

sure

men

t of

wat

er q

ualit

y

Imag

e dat

a bef

ore-

proc

essin

gin

land

mod

ule

Imag

e dat

a pre

-pro

cess

ing

in se

a mod

ule

Mul

ti-im

age

data

fusio

n




Database

System

txtALTOVA

HJ1A-CCD1-456-90-20101108-L20000855826JPG

HJ1A-CCD1-456-90-20101108-L20000855826XML

HJ1A-CCD1-456-90-20101108-L20000855826-1TIF

HJ1A-CCD1-456-90-20101108-L20000855826-2TIF

HJ1A-CCD1-456-90-20101108-L20000855826-3TIF

HJ1A-CCD1-456-90-20101108-L20000855826-4TIF

HJ1A-CCD1-456-90-20101108-L20000855826-SatAn

gletxt

HJ1A-CCD1-456-90-20101108-L20000855826-THU

MBJPG



















Acknowledgments


References





















































2rdquo Web






















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in




Knowledge phase






Geometriccorrection


Task



example

ModelsWeb services



the task



process chains

Service broker



Find

Find Register

Register

Binding

Binding

Geospatialknowledge

Servicerepository


Model broker

Service chain



Result

Execution engine









11198722 119872





GIS service

RS service

Compositionservice


onWSDL

Webservice


Webservice

interfaces(models

inputs andoutputs)



1

11987221198723 and119872











IName1 IName2




IName1

IName3

IName2

1lowast1

1lowast1



ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo




IName3Choose

IName2

ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo






Web service

Coordinatesystem



Web service


Web service

IName1


IName2

SOAP

Web service

SOAP

Web service

SOAP

Web service

SOAP

Web service










119899



SCHEME1 SCHEME






















Application layer

Logical layer


Representationlayer

Representation

Link

Associate

Function modules


Portal

Users

Project title

Map tool bar

Map

Function menus

Function panel


Representation

Link

ArcGIS



Remote sensing data

ASPNET Silverlight


Developed tool

Web service

ArcGIS server

ArcTools


Internet

server






River flows







Statistical report








Image result fusion




Image data fusion







data




Transparency




Rem

ote s

ensin

g m

onito

ring

syste

m o

f wat

er en

viro

nmen

t

Mea

sure

men

t of

wat

er q

ualit

yM

easu

rem

ent o

fw

ater

qua

lity

Mea

sure

men

t of

wat

er q

ualit

y

Imag

e dat

a bef

ore-

proc

essin

gin

land

mod

ule

Imag

e dat

a pre

-pro

cess

ing

in se

a mod

ule

Mul

ti-im

age

data

fusio

n




Database

System

txtALTOVA

HJ1A-CCD1-456-90-20101108-L20000855826JPG

HJ1A-CCD1-456-90-20101108-L20000855826XML

HJ1A-CCD1-456-90-20101108-L20000855826-1TIF

HJ1A-CCD1-456-90-20101108-L20000855826-2TIF

HJ1A-CCD1-456-90-20101108-L20000855826-3TIF

HJ1A-CCD1-456-90-20101108-L20000855826-4TIF

HJ1A-CCD1-456-90-20101108-L20000855826-SatAn

gletxt

HJ1A-CCD1-456-90-20101108-L20000855826-THU

MBJPG



















Acknowledgments


References





















































2rdquo Web






















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





GIS service

RS service

Compositionservice


onWSDL

Webservice


Webservice

interfaces(models

inputs andoutputs)



1

11987221198723 and119872











IName1 IName2




IName1

IName3

IName2

1lowast1

1lowast1



ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo




IName3Choose

IName2

ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo






Web service

Coordinatesystem



Web service


Web service

IName1


IName2

SOAP

Web service

SOAP

Web service

SOAP

Web service

SOAP

Web service










119899



SCHEME1 SCHEME






















Application layer

Logical layer


Representationlayer

Representation

Link

Associate

Function modules


Portal

Users

Project title

Map tool bar

Map

Function menus

Function panel


Representation

Link

ArcGIS



Remote sensing data

ASPNET Silverlight


Developed tool

Web service

ArcGIS server

ArcTools


Internet

server






River flows







Statistical report








Image result fusion




Image data fusion







data




Transparency




Rem

ote s

ensin

g m

onito

ring

syste

m o

f wat

er en

viro

nmen

t

Mea

sure

men

t of

wat

er q

ualit

yM

easu

rem

ent o

fw

ater

qua

lity

Mea

sure

men

t of

wat

er q

ualit

y

Imag

e dat

a bef

ore-

proc

essin

gin

land

mod

ule

Imag

e dat

a pre

-pro

cess

ing

in se

a mod

ule

Mul

ti-im

age

data

fusio

n




Database

System

txtALTOVA

HJ1A-CCD1-456-90-20101108-L20000855826JPG

HJ1A-CCD1-456-90-20101108-L20000855826XML

HJ1A-CCD1-456-90-20101108-L20000855826-1TIF

HJ1A-CCD1-456-90-20101108-L20000855826-2TIF

HJ1A-CCD1-456-90-20101108-L20000855826-3TIF

HJ1A-CCD1-456-90-20101108-L20000855826-4TIF

HJ1A-CCD1-456-90-20101108-L20000855826-SatAn

gletxt

HJ1A-CCD1-456-90-20101108-L20000855826-THU

MBJPG



















Acknowledgments


References





















































2rdquo Web






















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in






IName1 IName2




IName1

IName3

IName2

1lowast1

1lowast1



ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo




IName3Choose

IName2

ldquoInterfacerdquo

ldquoInterfacerdquo

ldquoInterfacerdquo






Web service

Coordinatesystem



Web service


Web service

IName1


IName2

SOAP

Web service

SOAP

Web service

SOAP

Web service

SOAP

Web service










119899



SCHEME1 SCHEME






















Application layer

Logical layer


Representationlayer

Representation

Link

Associate

Function modules


Portal

Users

Project title

Map tool bar

Map

Function menus

Function panel


Representation

Link

ArcGIS



Remote sensing data

ASPNET Silverlight


Developed tool

Web service

ArcGIS server

ArcTools


Internet

server






River flows







Statistical report








Image result fusion




Image data fusion







data




Transparency




Rem

ote s

ensin

g m

onito

ring

syste

m o

f wat

er en

viro

nmen

t

Mea

sure

men

t of

wat

er q

ualit

yM

easu

rem

ent o

fw

ater

qua

lity

Mea

sure

men

t of

wat

er q

ualit

y

Imag

e dat

a bef

ore-

proc

essin

gin

land

mod

ule

Imag

e dat

a pre

-pro

cess

ing

in se

a mod

ule

Mul

ti-im

age

data

fusio

n




Database

System

txtALTOVA

HJ1A-CCD1-456-90-20101108-L20000855826JPG

HJ1A-CCD1-456-90-20101108-L20000855826XML

HJ1A-CCD1-456-90-20101108-L20000855826-1TIF

HJ1A-CCD1-456-90-20101108-L20000855826-2TIF

HJ1A-CCD1-456-90-20101108-L20000855826-3TIF

HJ1A-CCD1-456-90-20101108-L20000855826-4TIF

HJ1A-CCD1-456-90-20101108-L20000855826-SatAn

gletxt

HJ1A-CCD1-456-90-20101108-L20000855826-THU

MBJPG



















Acknowledgments


References





















































2rdquo Web






















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in




SCHEME1 SCHEME






















Application layer

Logical layer


Representationlayer

Representation

Link

Associate

Function modules


Portal

Users

Project title

Map tool bar

Map

Function menus

Function panel


Representation

Link

ArcGIS



Remote sensing data

ASPNET Silverlight


Developed tool

Web service

ArcGIS server

ArcTools


Internet

server






River flows







Statistical report








Image result fusion




Image data fusion







data




Transparency




Rem

ote s

ensin

g m

onito

ring

syste

m o

f wat

er en

viro

nmen

t

Mea

sure

men

t of

wat

er q

ualit

yM

easu

rem

ent o

fw

ater

qua

lity

Mea

sure

men

t of

wat

er q

ualit

y

Imag

e dat

a bef

ore-

proc

essin

gin

land

mod

ule

Imag

e dat

a pre

-pro

cess

ing

in se

a mod

ule

Mul

ti-im

age

data

fusio

n




Database

System

txtALTOVA

HJ1A-CCD1-456-90-20101108-L20000855826JPG

HJ1A-CCD1-456-90-20101108-L20000855826XML

HJ1A-CCD1-456-90-20101108-L20000855826-1TIF

HJ1A-CCD1-456-90-20101108-L20000855826-2TIF

HJ1A-CCD1-456-90-20101108-L20000855826-3TIF

HJ1A-CCD1-456-90-20101108-L20000855826-4TIF

HJ1A-CCD1-456-90-20101108-L20000855826-SatAn

gletxt

HJ1A-CCD1-456-90-20101108-L20000855826-THU

MBJPG



















Acknowledgments


References





















































2rdquo Web






















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in




Application layer

Logical layer


Representationlayer

Representation

Link

Associate

Function modules


Portal

Users

Project title

Map tool bar

Map

Function menus

Function panel


Representation

Link

ArcGIS



Remote sensing data

ASPNET Silverlight


Developed tool

Web service

ArcGIS server

ArcTools


Internet

server






River flows







Statistical report








Image result fusion




Image data fusion







data




Transparency




Rem

ote s

ensin

g m

onito

ring

syste

m o

f wat

er en

viro

nmen

t

Mea

sure

men

t of

wat

er q

ualit

yM

easu

rem

ent o

fw

ater

qua

lity

Mea

sure

men

t of

wat

er q

ualit

y

Imag

e dat

a bef

ore-

proc

essin

gin

land

mod

ule

Imag

e dat

a pre

-pro

cess

ing

in se

a mod

ule

Mul

ti-im

age

data

fusio

n




Database

System

txtALTOVA

HJ1A-CCD1-456-90-20101108-L20000855826JPG

HJ1A-CCD1-456-90-20101108-L20000855826XML

HJ1A-CCD1-456-90-20101108-L20000855826-1TIF

HJ1A-CCD1-456-90-20101108-L20000855826-2TIF

HJ1A-CCD1-456-90-20101108-L20000855826-3TIF

HJ1A-CCD1-456-90-20101108-L20000855826-4TIF

HJ1A-CCD1-456-90-20101108-L20000855826-SatAn

gletxt

HJ1A-CCD1-456-90-20101108-L20000855826-THU

MBJPG



















Acknowledgments


References





















































2rdquo Web






















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





Database

System

txtALTOVA

HJ1A-CCD1-456-90-20101108-L20000855826JPG

HJ1A-CCD1-456-90-20101108-L20000855826XML

HJ1A-CCD1-456-90-20101108-L20000855826-1TIF

HJ1A-CCD1-456-90-20101108-L20000855826-2TIF

HJ1A-CCD1-456-90-20101108-L20000855826-3TIF

HJ1A-CCD1-456-90-20101108-L20000855826-4TIF

HJ1A-CCD1-456-90-20101108-L20000855826-SatAn

gletxt

HJ1A-CCD1-456-90-20101108-L20000855826-THU

MBJPG



















Acknowledgments


References





















































2rdquo Web






















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in
















Acknowledgments


References





















































2rdquo Web






















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in









































2rdquo Web






















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in



research article a framework for sharing and integrating remote sensing and gis...

Documents