System and method for automatically identifying locations in video content for inserting advertisement breaks
An automated method is provided for identifying candidate locations in video content for inserting advertisement (ad) breaks. Each candidate location is a different offset time from the beginning of the video content. Different distinct characteristics of the video content are identified at offset times. Certain characteristics are desirable and certain other characteristics are not desirable. Candidate locations are identified which have the most desirable characteristics at particular offset times, but which do not have any of the undesirable characteristics at any of the offset times.
1. An automated method for identifying candidate locations in video content for inserting advertisement (ad) breaks, wherein each candidate location is a different offset time from the beginning of the video content, the method comprising:
(a) storing in a memory one or more positive markers within the video content, each positive marker identifying a distinct characteristic of the video content, and defining one or more negative markers within the video content, each negative marker also identifying a distinct characteristic of the video content, each of the one or more positive and negative markers identifying different distinct characteristics of the video content, each positive and negative marker including one or more start and end locations within the video content that the characteristic appears, each start and end location being a time window where an ad break may potentially be started from;
(b) creating, using a video replicator, n number of identical copies of the video content, wherein n is equal to the total number of positive and negative markers, and associating each of the copies of the video content with a respective positive or negative marker;
(c) analyzing each copy of the video content, using a video processor, for the presence of the distinct characteristic associated with the respective positive or negative marker, and identifying all start and end locations within the video content where the distinct characteristic appears; and
(d) filtering out, by a filter, all start and end locations identified within the video content for the one or more positive markers where a distinct characteristic of one or more of the negative markers is identified within the same start and end locations, thereby deleting any time windows within such start and end locations as being time windows where an ad break may potentially be started from,
wherein candidate locations in the video content for inserting ad breaks are identified as being within the time windows of the start and end locations of the one or more positive markers that were not deleted.
2. The method of claim 1 further comprising:
(e) ranking, by an aggregator, candidate locations in the video content for ad break insertion based on a presence of positive markers at each offset time as determined from the start and end locations of the distinct characteristics of the respective positive markers, the ranking being computed based on a sum of weightings from each positive marker at a respective offset,
wherein the highest rankings represent the best candidate locations for ad break insertion.
3. The method of claim 2 wherein each positive marker further includes a confidence value for each start and end location within the video content that the characteristic appears, and wherein the weighting for each positive marker at the respective offset is multiplied by the confidence value,
wherein the confidence value indicates how confident is it that the characteristic of the positive marker is actually present between a particular start and end location.
4. The method of claim 3 wherein each positive marker further includes an importance factor, and wherein the weighting for each positive marker at the respective offset is further multiplied by the importance factor.
5. The method of claim 3 further comprising:
(e) further filtering out, by the filter, all start and end locations identified within the video content where no ad breaks should be inserted, thereby further deleting any time windows within such start and end locations as being time windows where an ad break may potentially be started from,
wherein candidate locations in the video content for inserting ad breaks are identified as being within the time windows of the start and end locations of the one or more positive markers that were not deleted, and also being within the time windows of the start and end locations that were not deleted as being time windows where no ad breaks should be inserted.
6. The method of claim 1 wherein the one or more negative markers include at least one of song playing and stunt scenes.
7. The method of claim 1 wherein the one or more positive markers include at least one of black frames, silent frames, scene change frames, and mood change frames.
8. The method of claim 1 wherein the distinct characteristics of the video content include audio and video characteristics.
9. The method of claim 1 wherein the video processor includes one or more of an image frame processor, an audio processor, and a caption analyzer, and wherein each copy of the video content is analyzed using one or more of the image frame processor, the audio processor, and the caption analyzer.
10. An apparatus for identifying candidate locations in video content for inserting advertisement (ad) breaks, wherein each candidate location is a different offset time from the beginning of the video content, the apparatus comprising:
(a) a memory that stores one or more positive markers within the video content, each positive marker identifying a distinct characteristic of the video content, and defining one or more negative markers within the video content, each negative marker also identifying a distinct characteristic of the video content, each of the one or more positive and negative markers identifying different distinct characteristics of the video content, each positive and negative marker including one or more start and end locations within the video content that the characteristic appears, each start and end location being a time window where an ad break may potentially be started from;
(b) a video replicator configured to create n number of identical copies of the video content, wherein n is equal to the total number of positive and negative markers, and associating each of the copies of the video content with a respective positive or negative marker;
(c) a video processor configured to analyze each copy of the video content for the presence of the distinct characteristic associated with the respective positive or negative marker, and identifying all start and end locations within the video content where the distinct characteristic appears; and
(d) a filter configured to filter out all start and end locations identified within the video content for the one or more positive markers where a distinct characteristic of one or more of the negative markers is identified within the same start and end locations, thereby deleting any time windows within such start and end locations as being time windows where an ad break may potentially be started from,
wherein candidate locations in the video content for inserting ad breaks are identified as being within the time windows of the start and end locations of the one or more positive markers that were not deleted.
11. The apparatus of claim 10 further comprising:
(e) an aggregator configured to rank candidate locations in the video content for ad break insertion based on a presence of positive markers at each offset time as determined from the start and end locations of the distinct characteristics of the respective positive markers, the ranking being computed based on a sum of weightings from each positive marker at a respective offset,
wherein the highest rankings represent the best candidate locations for ad break insertion.
12. The apparatus of claim 11 wherein each positive marker further includes a confidence value for each start and end location within the video content that the characteristic appears, and wherein the weighting for each positive marker at the respective offset is multiplied by the confidence value,
wherein the confidence value indicates how confident is it that the characteristic of the positive marker is actually present between a particular start and end location.
13. The apparatus of claim 12 wherein each positive marker further includes an importance factor, and wherein the weighting for each positive marker at the respective offset is further multiplied by the importance factor.
14. The apparatus of claim 12 wherein the filter is further configured to filter out all start and end locations identified within the video content where no ad breaks should be inserted, thereby further deleting any time windows within such start and end locations as being time windows where an ad break may potentially be started from,
wherein candidate locations in the video content for inserting ad breaks are identified as being within the time windows of the start and end locations of the one or more positive markers that were not deleted, and also being within the time windows of the start and end locations that were not deleted as being time windows where no ad breaks should be inserted.
15. The apparatus of claim 10 wherein the one or more negative markers include at least one of song playing and stunt scenes.
16. The apparatus of claim 10 wherein the one or more positive markers include at least one of black frames, silent frames, scene change frames, and mood change frames.
17. The apparatus of claim 10 wherein the distinct characteristics of the video content include audio and video characteristics.
18. The apparatus of claim 10 wherein the video processor includes one or more of an image frame processor, an audio processor, and a caption analyzer, and wherein each copy of the video content is analyzed using one or more of the image frame processor, the audio processor, and the caption analyzer.