Detecting of graphical objects to identify video demarcations
Particular embodiments analyze logos found in a video program to determine video demarcations in the video program. For example, a video demarcation may be content that marks (“marker content”) a transition from a first video content type to a second video content type. Marker content may be used so the user knows that a transition is occurring. Particular embodiments analyze the logos found in a video program to determine the video demarcations in the video. The video is first analyzed to determine logos in the video program. Once these logos are determined, particular embodiments may re-analyze the video program to identify marker frames that include the marker content that signal the transitions to a different video content types. The marker frames may be determined without any prior knowledge of the marker content. Then, particular embodiments may use the marker frames to determine video segments.
1. A method for identifying video demarcations in a video, the method comprising:
a first analysis by a logo template processor communicatively coupled to a storage device storing the video, the first analysis including:
i) analyzing the entire video;
ii) detecting one or more logos in the video;
iii) generating at least one logo template of the one or more logos that are detected, the logo template including a logo template image;
a second analysis by a marker frame processor, the second analysis performed using the at least one logo template, the marker frame processor communicatively coupled to the storage device storing the video, the second analysis including:
a) retrieving the video from the storage;
b) receiving the logo template from the logo template processor;
c) comparing the pixels in the logo template image against frames of the video to determine frames that may include the logo template image; generating an error signal indicating a matching error for every comparison of the logo template against a frame;
d) classifying frames of the video according to the presence or absence of the logo in the frame by comparing the values of the error signal to a threshold;
e) organizing the frames into frame sequences according to the classification;
f) classifying the frame sequences into multiple categories according to the presence or absence of the logo in the frame sequence;
g) identifying marker frames in the video that include marker content based on the comparison of pixels in the logo template image against the frames of the video, the marker frames being frames of the video that have an error signal below the threshold, wherein marker frames identify video demarcations that indicate transitions in the video from a first content type to a second content type and the marker frames are identified by matching frames at the boundaries of frame sequences of the same category;
h) outputting, to a video service processor, identifiers for the marker frames, wherein the identifiers indicate at least one of a frame number, a time in the video in which the marker frames are encountered, or other identification information for the marker frames; and
determining, by the video service processor, segments of different types of video content in the video using the marker frames.
2. The method of claim 1 , wherein using a pattern comprises comparing the logo in the set of logos to frames in the video to identify first frames that do not include the logo and second frames that do include the logo.
3. The method of claim 1 , wherein the organizing comprises:
classifying the first frames or the second frames into a first category of frame sequences and a second category of frame sequences, wherein the first category of frame sequences have a length shorter than the second category of frame sequences.
4. The method of claim 3 , wherein the first category of frame sequences are included in the plurality of frame sequences and the second category of frame sequences are not included in the plurality of frame sequences.
5. The method of claim 4 , wherein frame sequences in the second category of frame sequences have the length that is above a threshold.
6. The method of claim 1 , wherein determining the marker content comprises:
performing a cross correlation process using video content in the plurality of frame sequences to determine the marker content without prior knowledge of the marker content.
7. The method of claim 1 , wherein determining marker content comprises:
analyzing a first frame sequence in the plurality of frame sequences to determine a first signature from video content in at least a portion of the first frame sequence, the first signature representing the video content in the at least a portion of the first frame sequence; and
comparing the first signature with video content in other frame sequences in the plurality of frame sequences to determine the marker content.
8. The method of claim 7 , wherein the first frame sequence is randomly selected.
9. The method of claim 7 , wherein comparing the first signature with the video content in other frame sequences comprises:
determining second signatures from the video content in the other frame sequences; and
comparing the first signature with the second signatures to determine which second signatures match the first signature.
10. The method of claim 9 , wherein:
the first signature is from a first number of frames in the first frame sequence, and
the second signatures are from a second number of frames in the other frame sequences.
11. The method of claim 10 , wherein at least a portion of the first number of frames and the second number of frames are at a beginning of the first frame sequence and the other frame sequences.
12. The method of claim 9 , wherein comparing comprises:
determining a threshold for a match between the first signature and the second signatures;
generating an error match value for each comparison of the first signature with each of the second signatures; and
forming the plurality of frame sequences with frame sequences that have error match values that indicate the other frame sequences match the first signature based on the threshold.
13. A system for identifying video demarcations in a video, the system comprising:
a storage device storing the video;
a logo processor arranged to determine one or more logos, wherein the one or more logos are detected by analyzing a video;
a marker frame processor arranged to select a logo in the one or more logos;
a logo template processor communicatively coupled to the storage device storing the video, the logo template processor arranged to perform a first analysis, the first analysis including:
i) analyzing the entire video;
ii) detecting one or more logos in the video;
iii) generating at least one logo template of the one or more logos that are detected, the logo template including a logo template image;
the marker frame processor further arranged to determine a pattern associated with the logo in the video to identify a plurality of frame sequences in the video, wherein the pattern comprises a presence or absence of the logo in the video;
the marker frame processor further arranged to determine marker content for a set of marker frames by analyzing video content of the plurality of frame sequences, wherein the marker content is used to transition from a first content type to a second content type in at least of portion of the plurality of frame sequences in the video; and
the marker frame processor further arranged to determine a set of video demarcations based on the set of marker frames, the set of video demarcations used to segment the video into video segments
the marker frame processor further arranged to perform a second analysis, the second analysis performed using the at least one logo template, the marker frame processor communicatively coupled to the storage device storing the video, the second analysis including:
a) retrieving the video from storage;
b) receiving the logo template from the logo template processor;
c) comparing the pixels in the logo template image against frames of the video to determine frames that may include the logo template image; generating an error signal indicating a matching error for every comparison of the logo template against a frame;
d) classifying frames of the video according to the presence or absence of the logo in the frame by comparing the values of the error signal to a threshold;
e) organizing the frames into frame sequences according to the classification;
f) classifying the frame sequences into multiple categories according to the presence or absence of the logo in the frame sequence;
g) identifying marker frames in the video that include marker content based on the comparison of pixels in the logo template image, the marker frames being frames of the video that have an error signal below the threshold, wherein marker frames identify video demarcations that indicate transitions in the video from a first content type to a second content type and they are identified by matching frames at the boundaries of frame sequences of the same category;
h) outputting, to a video service processor, identifiers for the marker frames, wherein the identifiers indicate at least one of a frame number, a time in the video in which the marker frames are encountered, or other identification information for the marker frames; and
a video service processor arranged to determine segments of different types of video content in the video using the marker frames.
14. The system of claim 13 , wherein the processed video comprises only the at least the portion of the frame sequences.
15. The system of claim 13 , further comprising:
a video delivery system configured for sending the processed video to a client.