Broadcasting corporations and video archive operators usually use material from multiple sources (news agencies, private users, their own archives, etc.). That material is then edited by means of cuts, insertions/fade-ins, or the replacement of audio tracks. Identifying the broadcasting formats in which content is used and the channels through which it is distributed becomes increasingly difficult this way.

Video Segment Matching

To counter this problem, Fraunhofer IDMT has developed the Video Segment Matching, a software which recognizes identical video segments in different videos (even if content has been edited or transcoded), and determines their exact position and length. The tool extracts fingerprints from the videos under analysis, which are then matched against each other in a multi-stage, automated process. Users of the software benefit e. g. from more accurate accounting of copyright protected content, the detection of duplicates in archives, and improved management and storage of data.

Regarding the QC Components offered by the A/V Analyzing Toolbox, the Fraunhofer experts will be presenting new and enhanced functionalities, allowing automatic detection of errors and quality issues in digital audio and video content.

Music Detection

In addition, the A/V Analyzing Toolbox includes a component for automatic measurement of the exact amount of music in radio and TV programs. Copyright agencies may use this tool for accurate accounting and to make more efficient queries of programs, as music segments can be filtered out before concrete pieces of music are searched for by title.

Visitors interested in the A/V Analyzing Toolbox may feel free to stop by the Fraunhofer booth SU3714 in the South Upper Hall.