Temporal alignment of video recordings
AssigneeBruce Sharpe
InventorsBruce Sharpe, Malcolm Sharpe
Life of this patent
Estimating the term…
Abstract
Methods and apparatus are provided to establish temporal alignment of media clips. In an example embodiment, first and second media clips each contain an audio portion and the method comprises: determining an estimated global offset between the first and second clips; choosing a first test region of the first clip and identifying a corresponding second test region in the second clip based at least in part on the estimated global offset. The first and second test regions are compared to determine a local offset.
Claims20
A method for temporally aligning a plurality of video recordings of an event made by a plurality of cameras, the video recordings each including a video track recorded by one of the cameras and an audio track recorded by the same one of the cameras, the audio tracks of the video recordings containing audio data representing recorded sounds of the event, the method comprising:
- determining a temporal relationship between a pair of the video recordings based at least in part on processing the audio data representing the recorded sounds to determine an offset between the video recordings of the pair for which a direct or derived measure of similarity between the audio data for the video recordings of the pair of video recordings is maximized.
A system comprising a data processor configured by software to determine temporal alignment of a plurality of video recordings of an event made by a plurality of cameras, the video recordings each including a video track recorded by one of the cameras and an audio track recorded by the same one of the cameras, the audio tracks of the video recordings containing audio data representing recorded sounds of the event, the data processor configured to determine a temporal relationship between a pair of the video recordings based at least in part on processing the audio data representing the recorded sounds to determine an offset between the video recordings of the pair for which a direct or derived measure of similarity between the audio data for the video recordings of the pair of video recordings is maximized.
Dependencies are inferred from the claim text.