Images and Database Technology - Essay Example

Images and database technology Visualization and systematization of information have always been among the major priorities of human activity. Information was collected by generations and there often came the question what is the most optimal way to store and process it. With the appearance of computers this problem was solved and computers which were primary used for only numerical processing have become the storage centers for terabytes of information and data of different nature. Graphical abilities of today's media and computers impress: digital imaging allows not only to store and process digital images but also to develop and design new graphical objects and solve different graphical modeling tasks. Digital images are widely used instead of traditional pictures, photographs, movies, etc. Modern database technology allows to process data and information of different nature which is very important in today's world which is built on information streams and data exchange. The research paper different aspects of database technology, existing models of database management and their advantages compared to traditional way of data processing. The research paper also makes a review of the most popular DBMS and their characteristic features. It also discusses different aspects of storing images in the databases. Database technology The primary purpose of computers was only number processing. Later the file systems were introduced for data processing and processing of different sorts of information (with a relatively simple data structure) which were supported only by system programs. The development and improvement of computer technologies allowed the evolution of automated information systems as they required additional resources such as larger volumes of RAM , bigger storage devices, etc. Information system is a program complex for a reliable data storing and processing with a relatively complicated structure as it should have a visual and comfortable user's interface. Information systems differ from the file system as they allow the following features: Data management and description of data in EMD Management of RAM buffers (physical location and data navigation) Processing of the logically ordered file set Systematization and data restoring after critical system errors Multi-user interface of work with data Reliability of data storing, protection of data and access control to data Database Management Systems (DBMS) are the applied informational systems for management and processing of structured information. Such systems began to appear in late 60s. Modern DBMS are data information systems which support dynamic information model of a complicated controlled object: they allow to input, output, refresh and process data and also allow collective access to data. "Today's DBMS do not require data duplication and don't require frequent reorganization of data files" (Ambrosio,1989) As Elmasri, R., and Navathe, S. state in their work Fundamentals of Database Systems , DBMS have the features which allow several users to solve different tasks using the single dynamic information: During the process of file creation the conditions of uniqueness are formed and data which will be inputted into the file has to satisfy these conditions; Data description in the program doesn't depend on the description of the database in general, but is coordinated with database features and opportunities; Data management as well all functions of storing, developing, inputting , outputting in the database and help-functions of the system are autonomous running processes which are connected to applied programs only by information. Database Management System executes generalized inquiry function while specific inquires are formulated by user; Data is relatively independent from the program which allows to change definition of data and structure of data without additional program compilation. (from Elmasri, 1994) According to T. Dyck, the most widely used components in DBMS are the following : Table. The main component of DBMS which is used for storage and systematization of different data Memo-fields of tables. Memo-fields may contain different kind of data which often referred as notes. Keyboard macros. Are used in order to fasten the work with text information Index files. Are used in order to fasten data search and in order to organize the connections between different tables. Database. Database unites several tables by indexes as well as local and remote data sets. Program files. Contain codes for the dialogue organization and management of all components of the application. Menu. Allows the option of visual method of data processing. Reports and modified documents. Allow to view selected data from tables and different text information. Screen forms. Are used for visual editing and searching. They contain different control elements. SQL queries. Allow to analyze data and graphically visualize generalized information from one or several tables. Project. Unites all programs, databases and separate tables, menus, forms and other components. On the base of the project the application file is built. Class libraries. Are used in order to provide high productivity of db project. (from Dyck, 2002) C. Date, writes that depending upon the type of established links between data of the language of definite DBMS there were developed "different models of logical data organization: hierarchical, network and relational" (Date, 1989). One of the key concepts in DBMS is a data model, which is " a collection of concepts that can be used to describe the structure of a database - provides the necessary means to achieve this much needed feature of data abstraction. " (from Database and Database management. Retrieved October 20, 2005 from the web site: http://www.peterindia.net/Database.html) In hierarchical model the ties between different components are given in the form of tree-graphs which are single-sided from the upper tops to the lower tops of the graph. Such organization makes access to information easier only in the case when the query is reflected in the structure of the tree. This limitation is absent in the network model which uses the connection type "from one element to other elements" so that the single object may participate in the arbitrary number of such ties. The practical side of the network model realization appeared to be very complicated as it required a lot of physical and memory resources of the computer as it slowed the process of data processing. The concept of relational model which represents a set of relations over the subjects of objects definition was introduced by E. Codd in 1970. (Codd, 1985) It is based on the mathematical concept of relations, so that data is organized into "tables and unique strings which represent the value of the single attribute" (Codd, 1996) In the majority of modern DBMS such as DB2, SQL Server, Oracle, Informix, Sybase, Progress and others relational model is used for logical organization of data. As Date to Date:"the main advantage of relational DBMS is the simplicity of data processing" (Date, 1999) B. Tore defines states that the main disadvantages of relational model are "the usage of wide size of physical and operation memory" and dependence of the processing speed upon the table size. (Tore, 2002) Relational DBMS are divided into two types: systems which require the initial code of the program to be completed and linked into machine codes (for example Oracle) and systems where program is executed in command mode which allows easily detect and repair the errors and also analyze the work of every command. These systems use interpreter as a translator instead of compilation. Such systems differ by high visualization and high efficiency of processing. In the 1980s a number of customers was not satisfied by the limitations of the relational DBMS model. So the new object-oriented model was introduced which allowed the database to be associated as an individual project, so that the information was stored in one place instead of the number of tables used for the relational model: "An Object-Oriented Database Management System (OODBMS) also integrates more easily with applications that have been written with an Object-Oriented programming language such as C++ or Java. " (from Database and Database management. Retrieved October 20, 2005 from the web site: http://www.peterindia.net/Database.html) But their efficiency is questioned as there is no standard approach to OODBMS and their compatibility is very low. So no wonder that today RDBMS remain to be the mot popular DMBS for a variety of tasks and projects. DBMS Fox Pro is considered to be the fastest system among the systems of interpreter type. It has convenient user's interface, it supports the language of structured queries (SQL), it can be used in multi-user and network modes, as well as be used in performing the import and export of data on different levels including Internet connection. After the new operation system Windows was introduced, different relational DBMS allow to use Windows control components on the hand with support of object-oriented programming. So modern concept of the DBNMS allows to use different components and nearly excludes the process of programming during the design of the database as "functional options of the DBMS may be widened by existing technology of Component Object Model or COM" (Elmasri,1994) . This concept allows to create direct interaction between objects independently from their location in different applications. Today the problem of correct choice of the DBMS is a very important as the demands in data processing and data management grow rapidly. Sometimes it's important to store and process gigabytes of data in enterprises and the choice of the DBMS type is vital. Different programming tools such as C++ and Java allow to develop and design personal DBMS for any particular use, but nevertheless such systems as Oracle and SQL server remain to be the most popular as they contain a number of different functions from typical data functions to very specific and optional. The reliability and performance of both systems is nearly the same but they diverge in the compatibility: "SQL Server 2000 only works on Windows-based platforms, including Windows 9x, Windows NT, Windows 2000 and Windows CE. In comparison with SQL Server 2000, Oracle 9i Database supports all known platforms, including Windows-based platforms, AIX-Based Systems, Compaq Tru64 UNIX, HP 9000 Series HP-UX, Linux Intel, Sun Solaris and so on" (Chigrik ,2003) Today's development of database technology impresses. New methods which are introduced to relational DBMS and to network DBMS allow to fasten the process of data processing: such technologies as genetic algorithms are becoming more and more spread in applied technologies. It allows to develop new concepts of data control and makes the administration process faster and more reliable. Evolution of storing data in different databases. Fast development of computer technologies and development of applied computer science changed the concept of computers. Since 1970's computers start to be used not only as machines for calculations but as a powerful tool in data processing and data storing. Development of database technologies and of software engineering allowed to use multiple data functions in databases which were not limited to simple search function with common data of letters and digits. New generation of software programs allowed to support data of different types of files on a screen as it became possible to include different subprograms with support functions into the main program product. The same "functions of using multiple data types in database technology" became available in 1980's (Ambrosio, 1988). It's obvious that different databases may contain different types of data records as different kind of information may be requested: textual, visual, sound, etc. The problem with storing images was one of the first applied problems in database engineering as images represent one of the most important type of non textual data stored in databases and in other types of programs. One of the first solutions to this problem was including images into the database and other software products which needed visualization in the form of binary code. Nevertheless, it was understood that such solution is not optimal as it requires a lot of extra disk space and the size of the program grows dramatically (compared to the size of initial program and size of image separately). Later, when the concept of structured programming was supplement by other concepts such as object-oriented and relational models it was proposed to include a pointer in the program code which would define the route to the location of the image so that the image will be executed after the program starts. This concept allowed to save disk space and made the process of programs development quicker and more convenient. (according to Khan, 2002) These two concepts are the most popular solutions in modern DBMS, besides the concept of using pointers is the fundamental concept of storing images in databases in such DBMS as SQL, Oracle and ASP.NET. Both concepts are realized in database tools of different programming languages(Visual basic, Delphi, C++, Java, etc.) But nevertheless a big number of DB administrators and developers prefer to use pointer technology instead of storing binary data inside of the DBs if there size is relatively small. Faisal Khan gives obvious reasons for such choice: "Getting binary data in and out of databases is expensive, expensive in terms of server resources and time it takes for it. But having said that it is very convenient and in some cases extremely important to save all kind of data at a central place e.g, an ad server." (Khan, 2002) The majority of modern multi-function developer's tools such as VB, C++, Delphi etc. have the same concept of storing images in the database and moreover have the same methods for the realization of this task: either ImageList control or Binary Large Object method. These programming languages are more flexible and multi-functional than traditional DBMS as they allow to include additional functions and methods to database administration. The main advantage of these tools is that they allow to create database systems which will be the most customized and which will be of smaller size as they contain the minimum number of useless functions and tools. That's why pointer method is the most preferred in such types of developer's tools. Understandably, both methods have advantages and disadvantages. ImageList control has limitations for the number of images, so it becomes unpreferable when used in databases of larger size. In the article Store Images in Your database Andy Rosebrock and Stan Schultes describe different aspects of storing images in databases designed in VB: "Visual Basic offers several ways to store images, such as using the ImageList control, but these methods require that you recompile the application, and they're generally restricted to a limited number of images. The best way to store images is in a database. You can do this in a couple ways: by storing the image as a Binary Large Object (BLOB) in a database field, or by simply storing a "pointer" to the file location on disk. " (Schultes, 2005) These authors note that storing images in the database as a binary data will influence the size of application, besides it requires new compilation of the project and at the same time storing image "pointer" has negative aspects as in case the file is corrupt, deleted or moved from the location marked in pointer it will fail to be displayed in the database. That's why a number of experts advice to store images in the form of binary data if the size of the database is big and if a lot of free memory is allowed to be used, as this method is more secure. Different authors say that in administration of big databases storing images in separate location "is not reasonable and is not secure" (Waterson, 2005). First of all files may be removed or deleted or simply corrupt. In most of DBMS there are options which allow to restore deleted or missed information and data, so it makes the fist method more reliable. But at the same time Rosebrock and Schultes conducted comparative analysis of storing images in databases in the form of binary data and in pointer, results show that the second method is the most acceptable for small enterprise databases and small web databases: " In the example described earlier, with 100 records of 50K images stored in BLOB fields, the database grew to more than 4 MB. The same database using file pointers instead was under 100K. In speed comparisons, the file pointer method is the winner, completing the test in five seconds. These advantages generally make file pointers the preferred method of saving images." (Schultes, 2005). As I mentioned above, in modern DBMS such as Oracle and SQL the most preferred type of storing image data is storing it as Binary Large Objects which is considered to be more convenient and reliable. These systems have a lot of options for administering of larger databases where the main priority is high speed and reliability but not the size of the database application. Brian Peasland describes the procedure of storing images in Oracle as follows: "To store your binary images in an Oracle database, you will need to create a column in your table defined with the BLOB datatype. BLOB stands for Binary Large Object. Images from scanners, mpeg files, movie files and so on can all be stored in the BLOB column. An example of creating a table with a BLOB column can be seen below: CREATE TABLE test_table ( id NUMBER, image BLOB);" (Peasland, 2005) According to Ghiondea Alexandru's article Storing Images into a Database and Waterson, K. the procedure of storing images in the Microsoft SQL server is the same. Tonny Patton describes the procedure of storing images in SQL server as follows: SQL server breaks images so that they fit the db pages, which slows the process of assembling application : "Images stored in SQL Server are returned to the application via the TDS protocol and the data access APIs. Again, these are not optimal for image processing and impose overhead that doesn't exist with a file." (Patton 2005) Making a conclusion it's important to note that the choice of storing image method depends upon different factors in DB administration; type of databases, productivity of computer platform, performance of the DB platform and other factors. The choice of storing image method may be motivated either by the availability of the free disk space or by the demands to the reliability and security of the database integrity. In case there is shortage of free space it's better to use different pointer methods, otherwise storing images inside of the database would be the most reliable. References: 1. Ambrosio, J.(1988). Databases reach out to text and images - DBMS: Text and Image Management Software Magazine, 28 2. Chigrik, A.(2003) The comparison of SQL Server 2000 with Oracle 9i. Retrieved October 20, 2005 from the web site: http://www.databasejournal.com/features/mssql/article.php/2170201 3. Codd, E. (1985). "Is Your DBMS Really Relational" and "Does Your DBMS Run By the Rules" ComputerWorld 4. Date, C. J. (1989). Handbook of Relational Database Design Addison Wesley 5. Date, C. J. (1999). There's Only One Relational Model! Retrieved October 20, 2005 from the web site: http://www.pgro.uk7.net/cjd6a.htm 6. Database management system (2005). In Wikipedia. Retrieved October 20, 2005 from the web site:http://en.wikipedia.org/wiki/Database_management_system 7. Dr. Codd's 12 rules for defining a fully relational database(1996). Retrieved October 20, 2005 from the web site: http://www.cis.ohio-state.edu/sgomori/570/coddsrules.html 8. Dyck, T.(2002). Server Databases Clash, Retrieved October 20, 2005 from the web site: http://www.eweek.com/article2/0,3959,293,00.asp 9. Elmasri, R., & Navathe, S. (1994). Fundamentals of Database Systems. 2nd ed. Redwood City, CA: The Benjamin/Cummings Publishing Co. pp. 283 - 285. 10. Ghiondea, A. (2004). Storing Images into a Database. Retrieved October 20, 2005 from the web site: www.c-sharpcorner.com/code/ 2004/oct/StoringImagesIntoDatabase.asp 11. Haryanto, H.(2001). Storing Images in Database. Retrieved October 20, 2005 from the web site: http://codewalkers.com/tutorialpdfs/tutorial35.pdf 12. Hashmi, N. (2001). In Recent Memory.Retrieved October 20, 2005 from the web site: www.intelligenterp.com 13. Khan, F. (2002). Inserting Images ( binary data ) into Database. Retrieved October 20, 2005 from the web site: http://www.stardeveloper.com/articles/display.htmlarticle=2001033101&page=1 14. McLellan, T. (1994). What Is an Oracle Relational Database Retrieved October 20, 2005 from the web site: http://www.islandnet.com/tmc/html/articles/orareln.htm 15. Patton, T. (2005). Enhance the administration interface when storing images in SQL Server. Retrieved October 20, 2005 from the web site: http://www.builderau.com.au/architect/dotnet/soa/Enhance_the_administration_interface_when_storing_images_in_SQL_Server/0,39024710,39208484,00.htmfeed=rss 16. Peasland, B. (2005). Store images in an Oracle databases. Retrieved October 20, 2005 from the web site: http://expertanswercenter.techtarget.com/eac/expertAnswer/0,295208,sid63_gci1085188,00.html 17. Rosebrock, A. Schultes, S. (2005). Store Images in Your Database. Retrieved October 20, 2005 from the web site: http://www.zdnetasia.com/techguide/storage/0,39045058,39246987,00.htm 18. Save An Image Into SQL Server 2000 Database (2005). Retrieved October 20, 2005 from the web site: http:www.codeproject.com/aspnet/PicManager.asp 19. Tore, B. (2002). Introduction to Relational Databases - Part 1: Theoretical Foundation. Retrieved October 20, 2005, from 15 Seconds web site: http://www.15seconds.com/issue/020522.htm 20. Waterson, K. (2005). Storing images in MySQL. Retrieved October 20, 2005 from the web site: http:www.phpriot.com/d/articles/database/images-in-mysql/ Read More