Binary large object
Binary Large Objects ( BLOBs ) are large binary data objects such as B. image or audio files . The term is often used in connection with databases .
Databases
For a database , BLOBs are not further structured objects or field data. Some databases allow the field types to store large amounts of data (almost complete files ) as field content.
Character Large Objects ( CLOB or Text ) are similar to BLOBs, but contain strings.
If the column of a table is identified as a BLOB in a relational database system , most database management systems save the concrete LOB value separately and only note a key or a reference to this LOB value within the table .
BLOBs are mapped in common database systems using different data types:
- MySQL
- up to 256 bytes TINYBLOB , up to 64 KiB BLOB , up to 16 MiB MEDIUMBLOB or up to 4 GiB LONGBLOB .
- PostgreSQL
- BYTEA or via OID (Object Identifier)
- Oracle
- Previously LONG RAW from version 8i BLOB for binary data, CLOB for text data with 8-byte coding and NCLOB for Unicode data (up to 4 GiB)
- DB2
- BLOB for binary data, CLOB for text data with 8-byte coding, DBLOB for Unicode data (up to 2 GiB)
- Microsoft SQL Server
- varbinary from SQL Server 2005, text , ntext
In the Java programming language there is also an interface with the name Blob
in the JDBC , which connects Java with database systems. This allows you to determine the length of a BLOB and also to access the value yourself.
BLOBs as API for file access
Objects for reading, sending and processing binary data can be created in the browser scripting language JavaScriptBlob
.
Free software and open source projects
In the context of open source , the aim is that all program parts can be generated from a publicly available source code. Occasionally, however, proprietary components, in particular device drivers , are included in such projects that are only provided and licensed in binary form and whose source code is not published. Such components are known as “BLOBs”. The inclusion of such "BLOBs" in open source projects is controversial. In some free software projects (such as Debian ), according to the definition of “free software” by the FSF , such proprietary programs or drivers are not integrated and are removed when they are discovered.
Image processing
In image processing, contiguous areas are referred to as BLOBs that differ in terms of properties, such as the brightness values, compared to the surroundings. These are often referred to as regions or binary objects. The corresponding pixels are often segmented using a threshold value method. BLOBs are often used to create features for specific areas. For example width, height, roundness, average gray value of the pixels and so on. These features can be used, for example, to check whether the contact surfaces of the sample image are complete. An incomplete contact would have a smaller area.
Individual evidence
- ↑ Specification of the File API by the W3C (World Wide Web Consortium)
- ↑ Prof. Dr.-Ing. Reiner Dudziak: Script for the lecture Technical Image Processing (p. 134). (No longer available online.) Archived from the original on April 6, 2018 ; accessed on April 5, 2018 . Info: The archive link was inserted automatically and has not yet been checked. Please check the original and archive link according to the instructions and then remove this notice.
Web links
-
Interface
java.sql.Blob
Java API at Oracle - Blob MySQL reference