Fingerprinting
Workflow

MediaSeeker Core Platform Integration

build your own website for free

Fingerprinting in the Workflow

The MediaSeeker Core Platform is a flexible, scalable video content identification engine. It consists of two primary functional blocks:

- The ivitec Fingerprinting Engine (FE)

- The ivitec Database Indexing Server (DIS) 

Media Seeker Core Platform Blocks
  1. The Fingerprinting Engine (FE) analyzes each relevant frame
    and notes its unique characteristics.
  2. The extracted data are combined into a unique fingerprint file which represents that frame or series of frames throughout the production and archive process.
  3. The Database Indexing Server (DIS) stores the
    unique fingerprints in a database. 
  4. These fingerprints can then cataloged, indexed,
    stored in RAM and then can be queried / searched.
  5. All functions can be access manually through the ivitec UI,
    or programmatically through API function calls.
  6. When parameters inputted into the DIS are matched,
    output may be delivered in a variety of formats. 

ivitec Reference Database Creation

Reference video libraries (stock footage, video archives, …) are usually stored in a CMS and when the assets are digitized, it is common to fingerprint during that process. In order to search the video archives, it is needed to build up a reference fingerprinting database:

1. Create a reference fingerprint with the ivitec Fingerprinting Engine

2. Store reference fingerprint in ivitec Database Indexing server and link fingerprint to metadata           (e.g. name and CMS ID)

The ivitec fingerprinting process may happen in parallel or in serial and is easily integrated to other corporate systems.

Reference Database Creation

Content Identification

The MediaSeeker Core Platform searches any video source (e.g. internet, TV, video archives, repositories) for appearance of known video content. To search the ivitec reference database
with query videos:

1. Create a query fingerprint with the ivitec Fingerprinting Engine

2. Search query fingerprint in reference database (Database Indexing Server)

3. Get results/matches from Database Indexing Server. The output may be     
    delivered in a variety of formats.


Content Identification

MediaSeeker Infrastructure

Media Seeker can be deployed on a computing cluster to provide 24x7 scalable, robust analytical services

MediaSeeker Infrastructure

API Integration

The ivitec MediaSeeker Core Platform (MCP) provides the basic functionality to extract finger-prints from media files, to manage a repository of fingerprints using catalogs, and to search for content in catalogs using query fingerprints. All of these functions are available to the user through a comprehensive Application Programming Interface (API).  

The MCP API is based on web services that can be used from different types of programming languages and development environments. Assuming basic knowledge concerning web services referencing and usage, this guide gives a quick outline on the MCP API and shows its usage in examples.

We will provide you sample code to access the API.

MediaSeeker API Integration

MediaSeeker System Architecture

MediaSeeker System Architecture

ivitec MediaSeeker Core Platform

MediaSeeker provides core video fingerprinting functionality

Adaptive Fingerprint Creation
- Video and Audio Fingerprints

Metadata Generation
- Shot Segmentation
- Subtitles Identification
- Scene Classification (Facial, Studio) 

Fingerprint Indexing 


Fingerprint Matching 
- Video Scene Matching 
- Frame Exact Matching 
- Audio Detail Comparison 
- Image Detail Comparison 

Ivitec GmbH

Lange Reihe 29
20099 Hamburg

+49 6151 8509 - 0