Understanding ISO/IEC 15938‑2:2004 – The MPEG‑7 Description Definition Language

Scope, Technical Requirements, and Compliance Considerations for Multimedia Content Description

ISO/IEC 15938‑2:2004, also adopted as CAN/CSA‑ISO/IEC 15938‑2‑04, defines the Description Definition Language (DDL) for the MPEG‑7 multimedia content description interface. The DDL provides a rich syntactic and semantic framework for describing audiovisual data, enabling interoperable annotation, search, and retrieval across diverse applications. This article examines the scope, core technical requirements, implementation highlights, and compliance aspects of this foundational standard.

Scope and Applications of ISO/IEC 15938‑2:2004

ISO/IEC 15938‑2 belongs to the MPEG‑7 suite (Multimedia Content Description Interface). The DDL is the language used to create MPEG‑7 Description Schemes (DS) and Descriptors (D). Its primary role is to provide a grammar based on W3C XML Schema (XSD), extended with specific MPEG‑7 constructs that capture temporal, spatial, and semantic aspects of audiovisual content. The standard covers:

  • Definition of new XML Schema types for multimedia metadata.
  • Extensions beyond standard XML Schema to handle links, arrays, and semantic constraints.
  • Rules for creating compliant MPEG‑7 description instances.
  • Structural elements for describing time, media locations, and user preferences.
Tip: MPEG‑7 DDL is often confused with the entire MPEG‑7 standard. Part 2 specifically defines the language used to write the schemas in Parts 3 through 11. Understanding DDL is essential for anyone implementing or extending MPEG‑7 tools.

Technical Requirements and Core Features

The DDL is an XML Schema based language with several critical extensions. It mandates support for the following features:

Schema Constructs for Multimedia

ConstructPurpose in MPEG‑7Example
ComplexTypeDefine reusable element structures for description schemesMediaTimeType
SimpleType with constraintsRestrict values (e.g., durations, identifiers)durationType (P**Y**M**DT**H**M**S)
Choice / sequence groupsOrdering of descriptors and subdescriptorsVideoSegment choice of shot, scene, etc.
Attribute groupsCommon attributes (e.g., id, href)DescriptorAttributes
Unique/Key/KeyrefCross‑document referential integrityReferencing media source URIs

Semantic Extensions Beyond XML Schema

Standard XML Schema cannot fully express multimedia relationships. Therefore, the DDL incorporates:

  • Link elements (e.g., MediaLocator) for referencing external temporal segments.
  • Mpeg7Type base type that enforces required attributes such as id and version.
  • Semantic constraints using annotation elements to embed application rules (e.g., temporal ordering of events).
  • Array and vector support via specific list types for time series and feature vectors.
Important: Any extension to MPEG‑7 that introduces new descriptor types must validate against the DDL schema. If the schema lacks explicit constraints, the description may be syntactically valid but semantically non‑compliant. Always use the mpeg7 namespace and follow the schema modularisation rules.

Implementation Highlights and Schema Design

Implementing a system that uses MPEG‑7 descriptions requires careful adherence to the DDL’s modular architecture. The standard defines a set of core schema modules that can be imported or extended. Key implementation considerations include:

Schema Modularity and Namespace Management

MPEG‑7 schema modules are organised under the urn:mpeg:mpeg7:schema:2004 namespace (or 2001 for earlier versions). For compliance, custom schemas must:

  • Use the official namespace for core types.
  • Import modules such as Mpeg7Base.xsd and DescriptionSchema.xsd.
  • Avoid redefining types already defined in the standard.
Warning: Many implementations mistakenly use only the standard XML Schema validator. The DDL requires additional runtime semantic checks (e.g., temporal relationship validation) that cannot be expressed solely in XSD. Plan for an extra validation layer.

Performance and Extensibility

The DDL is designed to be extensible. When creating new description tools (e.g., for a specific video analytics domain), best practices include:

  • Deriving from existing abstract types (DSType, DescriptorType).
  • Using substitution groups to allow use of derived elements in place of base ones.
  • Employing the @xsi:type attribute for polymorphic instances.

Below is a small illustrative XML fragment showing a compliant use of DDL constructs:

<VideoSegment xmlns="urn:mpeg:mpeg7:schema:2004" id="seg001"> <MediaTime> <MediaTimePoint>T00:01:30:300</MediaTimePoint> <MediaDuration>PT00:00:05:00</MediaDuration> </MediaTime> <TextAnnotation type="scene description"> <FreeText>Car chase begins</FreeText> </TextAnnotation> </VideoSegment>
Success: The DDL’s extensibility has allowed the multimedia community to create domain‑specific profiles (e.g., for medical imaging, broadcasting archives) while retaining full interoperability with base MPEG‑7 tools.

Compliance and Testing Considerations

For a product or system claiming compliance with ISO/IEC 15938‑2:2004, the following must be verified:

Conformance to the DDL Schema

  • All MPEG‑7 description instances must validate against the official W3C XML Schema documents published by ISO/IEC.
  • Namespace declarations must be correct; use of deprecated namespaces (e.g., 2001) may cause validation failures.

Semantic Constraint Enforcement

Even if an instance passes XSD validation, it may violate DDL semantic rules. Compliance testing should include:

  • Verification of referential integrity (e.g., @idref values resolve to existing elements).
  • Check of temporal consistency (e.g., end time after start time, valid duration formats).
  • Enforcement of cardinality rules not captured in XSD (e.g., exactly one MediaTime in a Segment).

Adoption Notes (CAN/CSA)

The Canadian adoption CAN/CSA‑ISO/IEC 15938‑2‑04 is identical to the international text. Products sold in Canada may reference the CAN/CSA version for regulatory acceptance. No additional national deviations exist.

Best Practice: Use existing conformance toolkits (e.g., MP7‑Suite, SAMBA) that include both schema and semantic checkers. These tools identify the majority of non‑compliant descriptions before integration.
Q: What is the difference between MPEG‑7 DDL and standard XML Schema?
A: The DDL extends XML Schema with multimedia‑specific constructs (temporal types, links, semantic annotations) and imposes rules on how these constructs must be combined to describe audiovisual content. Standard XSD validation alone does not enforce these semantic rules; a DDL‑aware parser is required.
Q: Is ISO/IEC 15938‑2:2004 still relevant today?
A: Yes. While newer standards (e.g., MPEG‑21, Ontology for Media Resources) have emerged, the MPEG‑7 DDL remains in use in broadcasting archives, digital preservation, and video analytics. Its extensibility allows integration with modern metadata formats.
Q: Can I use any XML Schema validator to check MPEG‑7 instances?
A: For syntactic validation, any XSD 1.0 processor works. However, semantic constraints (e.g., temporal consistency, unique keys) require additional checks. Always perform both syntactic and semantic validation recommended by the standard.

Published 2026 – Technical Reference Article

📥 Standard Documents Download

🔒
Please wait 10 seconds, the download links will appear after the ad loads

Leave a Reply

Your email address will not be published. Required fields are marked *