Digital content conversion quality control system and method
AssigneeNBCUniversal Media, LLC
InventorHarvey Landy
Life of this patent
Estimating the term…
Abstract
Automated, algorithmic quality control is performed for digital content converted from one form or format to another. Such conversion may be made of movies, television programs, feature films, advertisements, or any other content. The conversion process may be semi or fully automated, and may include a range of alterations, such as pulldown or frame rate conversions, size and/or resolution conversions, addition of content, deletion of content, and so forth. Actual content contained in pre- and post-converted frames is utilized, such as by image recognition techniques, as a base for the quality control routines. Audio data may also be considered. The quality control is fully or nearly fully automated with minimal manual involvement.
Claims11
A method comprising:
- for a digital video content file comprising successive frames of video content, converting, via a processor, the digital video content file from a first version at a first frame rate to a second version at a second frame rate that is different from the first frame rate;
- performing, via the processor, an automated quality control process by:
- identifying, via the processor, a frame in the first version;
- determining, via the processor, a corresponding frame in the second version that corresponds to the frame in the first version;
- identifying, via the processor, at least one subsequent frame in the first version;
- determining, via the processor, a corresponding subsequent frame in the second version that corresponds to the at least one subsequent frame in the first version;
- and
- verifying, via the processor, that the second version is properly converted by comparing at least one characteristic of the frame in the first version to at least one characteristic of the corresponding frame in the second version, and by comparing at least one characteristic of the at least one subsequent frame in the first version to at least one characteristic of the corresponding subsequent frame in the second version;
- wherein the conversion comprises converting frames in the first version to frames in the second version, wherein a first subset of the frames are combined in the second version, and a second subset of the frames are not combined.
Dependencies are inferred from the claim text.