A Evaluation of Integrating GIS databases with Photogrammetric Systems and its application...

A Evaluation of Integrating GIS databases with Photogrammetric Systems and its application in Urban Management

SHARE

S. Hosseinian
Graduate student, Dept of Geomatic Engineering
K.N.Toosi University of Tehran-Vali asr Street-Tehran-Iran
[email protected]

Dr. H. Ebadi
Assistant Professor, Dept of Geomatic Engineering
K.N.Toosi University of Tehran-Vali asr Street-Tehran-Iran
[email protected]

F. Farnood
Ph.D. student, Dept of Geomatic Engineering
K.N.Toosi University of Tehran-Vali asr Street-Tehran-Iran
[email protected]

ABSTRACT
Spatial Data play an effective role in different applications especially urban applications. GIS databases are used to manage the collected spatial data. Using these systems brings us benefits such as security, accuracy and possibility of data integration. It is struggled to use GIS databases and integrate them with spatial data acquisition systems especially photogrammetric systems that are important and reliable spatial data sources. Integration of photogrammetric systems and GIS databases makes it possible to control data quality, update and improvement of existing data and map production in various scales.

Currently, various systems in different levels of integration have been introduced to the market. So it is necessary to evaluate integration approaches and different kinds of integrated systems in order to use them in our applications. In this paper, after evaluating spatial databases and photogrammetric systems, integration systems are compared in various aspects such as efficiency, data quality and connection type. It presents the results of this comparison with considering their integration level. It is shown that by increasing the level of integration, the efficiency and data quality of the system is increased. Finally, with considering the results of this evaluation and the requirements of urban applications that are mentioned in this paper, it is recommended a new approach of integrating these systems for urban management.

1- Introduction:
Spatial database management systems have been presented to maintain and manage spatial data. Using such systems provides the possibility of integration of collected information from different sources, data updating and validation and map generation in various scales. Therefore, many attempts have been made to use these systems and integrate them with spatial data acquisition systems. One of the most important and common systems of spatial data acquisition systems particularly in urban applications is photogrammetric systems.

In recent years, integration of spatial database management systems and photogrammetric systems has been extensively taken into consideration and various integrated systems have been presented. By direct and precise integration of photogrammetric systems and spatial database management systems, all information gathered by photogrammetric methods is transferred directly to GIS databases with maintaining the precision. This provides the possibility of instant GIS analysis, and integration of collected data with data from other sources. This will assure geometrical, topological and semantic adoptability of data. This integration makes it possible to use information from GIS databases in the process of automatic or semi-automatic data acquisition and updating. Integration of photogrammetric systems and GIS databases also reduces time and costs of producing and maintaining high quality and update information in spatial databases. Regarding the above advantages for integration of photogrammetric and spatial database systems, these systems could be very effective and useful in urban applications and satisfy substantial needs for production, maintenance and organization of up-to-date and accurate spatial database. Therefore integration of these two systems is very important in urban applications.

Due to the importance of integration of these two systems it is necessary to make precise studies in this field. In order to evaluate integration of photogrammetric systems and spatial database it is first necessary to study different types of spatial database and common and important Photogrammetric Systems. For this purpose, chapters 2 and 3 of this paper have been allocated to this topic. Then in chapter 4 we have precisely studied integration of photogrammetric systems and spatial databases and its methods. We have presented examples of existing integrated systems and evaluated them and state their application in urban management. In chapter 5, we present conclusion and suggestions according to the topics provided in previous chapters.

2- Spatial Database Management Systems
Using spatial database management systems allows creating, saving, retrieving, editing, manipulating and analyzing spatial and non-spatial data in different GIS applications. These systems provide possibility for quick access to spatial data, spatial query, control data redundancy, maintain and control accuracy of spatial data and the relations among them and maintaining adoptability and control of data validity. In common spatial database management systems, topologic model is used to represent spatial data sets. This reduces the volume of required editing operations, besides it increases system precision and consistency to a great extent. Spatial database management systems have an extensive application of functions, and provided the possibility of maintaining, updating, changing, displaying, and controlling data. Capabilities of spatial database management systems have led to high efficiency of these systems for different affairs. Today a variety of database management systems have been introduced to the market [10].

Database systems are usually classified according to the data model they use. In the past, there were the hierarchical and network data models, which have now being replaced by three models, namely, relational, object-oriented and object-relational: (a) Relational databases are based on the relational model, which organizes the data into relations or tables [4]. (b) Object-oriented database management systems (OODBMS) integrate object orientation and database functionality [12]. An object-oriented database is characterized by having an object-oriented logical data model and by using an object-oriented programming language as its principal interface [4]. These systems are flexible but are used less because they don’t have a standard query language. ObjectStore, FastObject and Gothic database systems are also samples of object-oriented database. (c) Object-Relational approach is a compromise between the concepts of the object-oriented and relational models. In this approach, object-oriented features are incorporated into relational databases in a manner that utilizes the power of object orientation while maintaining the full functionality of the relational model [4]. Oracle and Informix are samples of objectrelational database systems.

In both of object-oriented databases and object-relational databases, data is modeled in the form of object. The most important spatial database systems, which are particularly used for integration with photogrammetry in GIS applications apply object-orientation in different ways[11]: Laser-Scan’s Gothic (Laser-Scan, 2001) uses an object-oriented database, GE Smallworld (GE SmallWorld, 2001) uses a relational database termed The Version Managed Data Store (VMDS) combined with an object-oriented application programming language, and Oracle (Oracle, 2001) uses an object-relational database with an abstract data type for spatial data. Each of these systems has particular characteristics that will be explained briefly in the following.

Oracle spatial is an object-relational database system. This system has high interoperability. Powerful functions and operators are available in this system. Using this system we can implement query on spatial and non-spatial data in the database. Here it is possible to maintain data security by controlling data accessibility. This system supports topologic structure. Also in its 10g version it is possible to work with raster data beside vector data. In order to access spatial data stored in the database, we can use Oracle spatial directly or via ArcSDE (ESRI) as a gateway.

Gothic object-oriented database from Laser Scan Company allows working with dynamic effects [6]. This system has various methods such as methods of data validation and displaying methods. Using these methods, data precision is maintained and the user could apply desired rules on each class of effects. In this way database would apply these rules on the effects at the time of entry or change. Displaying methods also allow dynamic display. Gothic database supports topologic structure. Gothic database can hold a mixture of vector objects and raster DTMs [7].

Smallworld database technology has been developed by GE (General Electric) Company. GE Smallworld uses relational database called VMDS (Version Managed Data Store) combined with applied object-oriented programming language. Smallworld allows different types of spatial database saved in a database and makes it easy to use those data. Thousands of users can use Smallworld to work with a high volume of data in complicated and distributed operations. If Smallworld functions couldn’t satisfy the application needs, it would not be possible to develop the system easily.

3- Photogrammetric Systems
Photogrammetric systems extract spatial data from images [10]. New generation of photogrammetric systems is DPWS (Digital Photogrammetry Workstations). DPWS is a hardware and software that extracts photogrammetric products from digital images.


Figure 1: LH System DPWS

Photogrammetric Systems differ in various aspects such as functionality, automation degree and price [10]. They can be classified based on degree of automation and functionality. In one group we can consider multi-purpose systems that have been presented by traditional providers such as LH Systems, Automatic and Z/I Imaging. Also some systems that have been presented by new providers are Inpho, Erdas, Partners, and Supresoft [10]. Using these systems, it would be possible to accomplish tasks such as image orientation, DTM generation, producing orthophotos and orthophoto mosaics and collecting vector data.

These systems have various visualization tools. Except identification and measurement of ground control points and collecting vector data that are usually made in a CAD environment, other tasks are automated. Stereo viewing is available in all these systems. The second group is systems with automation degree less than the first group and developed by smaller companies. Some of these companies have a longer history such as DVP Geomatis, ISM, KLT Associates, and R-Wel, and some of them have been presented just recently such as 3D Mapper, Espa Systems, TopoL Software/Atlas and Racurs. The third group is remote sensing systems such as ER Mapper, Matra, MicroImages, PCI Geomatics, and Research Systems, which are mostly used for production of orthophotos, mostly without Stereo viewing capability [10].

4- Evaluation of Integrating Spatial Database Management Systems and Photogrammetric Systems
By integration of spatial database management systems and photogrammetric systems, it would be possible to transfer all information gathered by photogrammetric method to spatial databases. So accuracy of data is maintained and it would be possible to integrate collected data with other data and direct data control. In this case validity and consistency are maintained and the volume of editing operations, as well as required time and costs are decreased. Using such systems, we can make decisions and planning very precisely.

Therefore, in applications where spatial data play an important role such as urban applications, using these systems could be very effective. With regard to the importance of these systems and their extensive use in urban applications it is necessary to evaluate different integration methods. Moreover in order to use integration systems in different applications it is necessary to study different types of systems so as to select a suitable system in each application or create new systems if required. We can classify methods of integration of photogrammetric systems and spatial databases based on various aspects such as connection type, kind of systems and efficiency of systems. Integrated systems can be classified into two major groups based on the criteria of the connection type between photogrammetric systems and spatial database, which influence on efficiency of the system: (a) First group: Integrated systems in which photogrammetric systems and spatial databases are connected to each other with an internal connection and for this purpose some tools have been considered inside these systems. Presenting such systems is limited only to providers of systems because necessary tools should be available in order to create relationship inside Photogrammetry and database systems.

Connection between Erdas Imagine with ArcInfo from ESRI, integration of Finnish ESPA with Smallworld and integration of PCI system with Oracle are samples of systems that are connected to each other by direct link inside the systems. In some of these systems such as Socet Set and Lamps2 developed by Laser Scan, this link is much more powerful than other systems and these systems have high efficiency. Samples of such systems will be explained later in this chapter. Since facilities of systems in this group have been provided by companies, they and they may not be completely consistent with user needs in different applications. (b)Second group: Integrated systems in which photogrammetric systems and spatial databases are connected to each other with an external connection. In this case, the important issue is defining an efficient interface between photogrammetric systems and spatial databases that can transfer data directly and provide possibility of active data validation. In this group, various kinds of photogrammetric and database systems can be connected to each other. By developing a powerful interface you can choose photogrammetric and database systems by considering the application needs.

These interfaces can present various facilities such as data structuring. They can be developed by programming languages like VB, C++. Some advantages of using this method for connecting photogrammetric and spatial database systems are consistency of interface software to the application, flexibility and possibility of choosing favorite kinds of photogrammetric and spatial database systems. But it should be considered that creating an efficient interface is not easy and it should be used in particular problems. After evaluating existing integrated systems if they couldn’t fulfill the application’s requirements, this method can be useful.

4-1- A Study on existing integrated Systems
We can use collected information from the images and directly put them into the database. For the first time Dowman (1990) characterized a photogrammetric workstation as being an active window into the 3D GIS database [10]. Afterward there have been many attempts in this regard and systems with different capabilities have been presented for this purpose. In order to compare current integrated systems, they are explained here: Lamps2/Socet set system: This system has been presented by Laserscan and is the integration of Socet Set photogrammetric system and LAMPS2 system exists. Socet set system is a photogrammetric system for extracting spatial data. LAMPS2 and the Leica Helava Socet-Set photogrammetric suite present a unified interface to the user. The Gothic database at the heart of LAMPS2 holds its model of the real world as objects that have properties and behaviors. In this system, height information has been held at the object level as height attributes [7]. This integration offers different facilities such as 3D position input in real-time from the stereo instrument direct into LAMPS2, Interactive digitizing and editing, direct into the object-oriented spatial database, Ergonomic cartographic editing tools for database update and Object-Oriented integrity, validity controls enforced and 3D vector stereo superimposition. It also uses the advantage of the of object oriented database of gothic.


Figure 2: Lamps2 / Socet Set system

PCI Geomatics system: PCI Geomatics software is provided by PCI Geomatics an image-centric geospatial solutions company. In PCI Geomatics, the possibility of working with oracle spatial and GeoRaster in the software is provided. PCI’s GeoRaster ETL for Oracle product is a gateway to Oracle 10g. GeoRaster ETL is an Extract, Transform and Load tool for Oracle 10g Spatial and GeoRaster [28]. GeoRaster ETL also includes a scripting environment for customized or command line loading and a visual workflow builder to build process flows which incorporate transformations in the loading process [28]. PCI Geomatics software extends capabilities of managing spatial and raster data in the database of oracle in different ways such as providing tools for uploading and downloading data into Oracle, Intuitive viewer for spatial data management and analysis and providing Cutting-edge functions for remote sensing, photogrammetry, GIS, and mapping.

Erdas Imagine system: IMAGINE Essentials is part of the ERDAS IMAGINE from Leica Geosystems GIS & Mapping, provides all of the fundamental tools for GIS professionals to produce quality projects. IMAGINE Essentials provides you with an intuitive and customizable interface to manage geographical information [26]. IMAGINE Essentials can stand alone or work alongside other GIS software packages, such as ESRI ArcGIS or MapInfo. IMAGINE Essentials was designed to process geographic image data quickly without compromising the integrity of the data. This software can read, create or edit geodatabase, SDE of simple vector and shape file [26]. Imagine essentials can be used in ERDAS Imagine environment for integration, updating, manipulating and coverage of ArcInfo. IMAGINE Essentials offers raster and vector-image manipulation and management, 2D-modeling tools with surface feature simulation, spectral image classification and map generation tools.


Figure 3: Erdas Imagine system

ESPA system: ESPA system from Finnish company incorporated collection and editing of topographic data on the basis of digital aerial photographs into Smallworld software [29]. Over the implemented stereo workstation data transfer link, features in the Smallworld database can be edited on the basis of stereo images. Espacity has tools for measuring, computation and visualizing data. Espacity uses polarization or anaglyph techniques for stereo viewing. By using EspaGate link, it would be possible to transfer spatial data between Espacity and other systems. By using EspaGate, Espacity can be linked to GIS, CAD and database systems such as AutoCAD, Microstation, Smallworld. Espacity can transfer collected data automatically to GIS, CAD and spatial database systems.

Z/I Imaging system: This photogrammetric system with the Intergraph Geomedia is another sample of integrated systems. This system is presented by Z/I Imaging provider as an effective system in the GIS and Photogrammetry world. This system can gather information from images and transfer them to GIS. Workstation of OrthoPro is one of the products of Z/I Imaging [15]. In OrthoPro workstation, users can implement different operations such as edit, rectification and Mosaic. An ortho project is a file of access database which includes all predefined features and other data used for developing OrthoPro products. Geomedia data is displayed and edited and saved for OrthoPro and integrate information and act as a displaying and analyzing tool and a platform for GIS. Geomedia can be used on MS access database or Oracle [15]. Geomedia allows the users to make questions about information and save the answers. Questions may include objects of some GIS database such as MGE, Arc/Info and ArcView. Geomedia does not need any special language but it can use tools such as visual basic and power builder.

4-2- Evaluation of Existing Integrated Systems
Considering the importance of integrated systems and their increasing usage in urban purposes, in addition to evaluating integration methods which presented in this section, it is necessary to evaluate common integrated systems explained in this section choose appropriate integrated system for our application and if it is needed develop a new integrated system. These systems are compared with respect to criteria such as degree of automation, efficiency, data quality, connection type, simplicity and kinds of their Photogrammetric and spatial database systems. Results of this comparison, shown that Lamps2/Socet set system is in higher level than other systems in aspects degree of automation, efficiency and data quality. Also connection of photogrammetric system and spatial database is closer in this integrated system than others. Since Lamps2/Socet set system have possibility of active manipulation, so quality of produced data in this system is higher than others. But Erdas Imagine system and PCI Geomatics system are simpler systems than Lamps2/Socet set system. Espa and Intergraphs are used more for cad application. Based on the results, each of this system can be useful for a particular application.

At the same time no system can be selected for all applications because each of them has advantages and disadvantages. For example Lamps2/Socet Set system is better in aspects such as validation, automation than other systems but there are other integrated systems that are better in aspects such as simplicity. Another important issue is its object-oriented database (Gothic) applied in this system. One of the systems that are simpler than Lamps2/Socet Set system is ERDAS Imagine. Therefore, it is impossible to choose a particular system as the best one for all applications and system needs should be taken into consideration before any selection. Selecting an integrated system is an important issue especially in urban applications. Different aspects of urban applications such as spatial data collection, up-to-date and precise map production, 3D city modeling and 3D GIS can be influenced by integrated system extensively. Updating and accuracy improvement, GIS databases by using integrated systems will be done more precisely and so faster. These are very important issues for various urban applications. In all urban applications we need precise and update maps. Another important issue in management of urban infrastructure is 3D city modeling and providing 3D GIS of cities. (As present urban planning and management is done based on 2D data, lack of 3D data faces these issues with problems and for solving these problems, it has been suggested that 3D city models and 3D GIS can be used.) Geographic information system, digital Photogrammetry and Remote Sensing are used techniques in modern urban infrastructure management. For solving existing problems in urban management, 3D city modeling and providing 3D GIS, it is recommended to develop a direct system with on-line and automatic 3D data structuring and validating tools with respect to application requirements, after evaluation of existing researches. From evaluating existing recent researches in this field, it is resulted that on-line validation using rule-based processing is needed to efficiently support the more complex data models coming into use (Woodsford, 2004). In this way, a rule-based system can be developed for on-line and automatic validating and applying consistency constraints on data in an object-oriented environment.

5- Conclusion and Suggestion
In this paper, after studying spatial database management systems and photogrammetric systems, considering importance of integration of these two systems, integration methods are evaluated and popular existing integrated systems were studied. Each of the presented methods for integration has characteristics that should be considered for selecting a suitable method for our needs. In the first method of integration, internal link, that is explained, we can use system facilities provided by photogrammetric and database system providers for integration but this has some restrictions and may not match to the application requirements. By using the second method of integration, external link, that is explained, requirements of the application are considered but in this way it is needed to develop a new link between systems with respect to requirements and this may not be so easy. Using this method it is possible to choose desired photogrammetric and database systems with considering the requirements of the application. After comparison and evaluation of integrated systems, it can be said that each of these systems has advantages and disadvantages.

So, with respect to their characteristics they can be useful in different applications due to application requirements. Considering the characteristics of integrated systems, we can use these systems in urban applications to maintain and control data accuracy and to update and improvement of existing data. In this way, high functionality, accuracy and consistency will be achieved. On the other hand, time and costs are considerably saved. So, integration of photogrammetric system and spatial databases can be very effective especially in precise urban management and planning. Since the controlled images are main sources for data improvement, and updating, the systems which collect data from the images and transfer them directly into the databases with maintaining data validity, play important roles in applications that need precise and upto- date data. Among these applications, we can mention to urban management and resource management.

Integration of capabilities of a GIS which uses a precise and up-todate database with hydrologic models makes precise data analyzing in water management possible. Another application of system integration is in urban infrastructure management as digital photogrammetric products are regarded as main information sources of urban infrastructure. Some important issues in modern urban infrastructure management are urban planning and management, 3D city modeling and 3D GIS. Currently, integrated systems explained in section 4 are applied for 3D city modeling and providing 3D GIS extensively. Therefore, using integrated systems for spatial data production and management, 3D city modeling and providing 3D GIS of cities affect on various aspects of urban management and planning. So these integrated systems are recommended to be used in urban management more extensively. For solving problems in urban management by considering recent researches it is suggested to develop and use a direct system with on-line and automatic 3D data structuring and validating tools using rule-based processing with respect to application requirements.

References

  1. Berry, K. (2000) GIS Technology in Environmental Management, Colorado USA.
  2. Carrico, M. A., J. P. Jones, J. C. Girard, (1989) Building knowledge systems, McGraw-Hill Book Co., New York.
  3. Cooper, R. (1997), Object databases: An ODMG approach. International Thompson Computer Press.
  4. Elmasri, R., Navathe, S., Database Management System, University of Texas, Georgia Institute of Technology.
  5. Hachler, T. (2003) Online Visualization of Spatial Data, University of Zurich.
  6. Hardy, P.G., Map Production From An Active Object Database, Using Dynamic Representation and Automated Generalisation, Laser-Scan Ltd, Cambridge.
  7. Hardy, P.G, Cameron, E.C.M., Stereo Images with Active Objects-Integrating Photogrammetry with an Object Database for Map Production, Laser-Scan Ltd, Cambridge.
  8. Hayles, M.,Integration of Photogrammetry and 3D Geo-spatial Databases, Laser Scan, Cambridge.
  9. Heipke, C. (2001) Digital Photogrammetric Workstations – A review of the state-of-the-art for topographic applications, University of Hannover.
  10. Heipke, C. (2002) Some Requirements for Geographic Information Systems: A Photogrammetric Point of View, ISPRS.
  11. Jakobsson, A., Salo-Merta , L., (2001) Definition of a Basic Topographic Framework for National GI Policy – One Database for All Basic Topographic Data, Helsinki University Of Technology.
  12. Jinmu, C., A Rule-Based Expert System Using an Interactive Question-and-Answer Sequence, University of Geogia.
  13. Khoshafian, S. (1993), Object-oriented databases, John Wiley & Sons.
  14. Li, D., Wu, H., On Developing Spatial Data Infrastructure of China, Wuhan University.
  15. Madani, M. (2001) Importance of Digital Photogrammetry for a Complete GIS, Global Data Infrastructure Conference Cartagena, Columbia.
  16. Madani, M. (1999) OrthoPro: The Z/I Imaging Approach for the Ortho Production, Photogrammetric Week ’99’.
  17. Murai, Sh. (1998) GIS Work Book, University of Tokyo.
  18. Nebiker, S., Relly, L., Concepts and system Architectures for the Management of Very Large Spatial Raster Objects in a Database Framework, Swiss Federal Institute of Technology Zurich.
  19. Negnevitsky, (2002) Rule-Based Expert Systems, Peerson Education.
  20. Riaux, PH., Scholl, M., Voisard, A. (2002) Spatial Databases with Application to GIS, Morgan Kaufmann Publishers, USA.
  21. Shekhar, S., Chawla, S. (2001) A Tour of Spatial Databases, University of Minesta.
  22. Spiller, R. (1999) The Z/I Imaging: A New for Photogrammetry and GIS, Photogrammetric Week ’99’.
  23. Twumasi, B. O.,(2002) Modelling spatial object behaviours in object-relational geodatabase.
  24. Wahl, R., The Use of Topology on Geologic Maps, U.S. Geological Survey. [25] Woodsford, P. A. (2004) System Architecture for Integrating GIS and Photogrammetric Data Acquisition, Laser-Scan, UK.
  25. (2004) Tools for GIS Professionals, IMAGINE Essentials Whitepaper.
  26. (2007) Leica Ortho Accelerator Product Description, Leica Geosystems.
  27. (2005) Using PCI Geomatics Software with Oracle 10g Spatial and GeoRaster, A PCI Geomatics Whitepaper.
  28. (2002) Newsletter, OEEPE No.2.