Method and apparatus for generating video fingerprint
AssigneeAlibaba Group Holding Limited
InventorChangguo Chen
Life of this patent
Estimating the term…
Abstract
A method and an apparatus for generating a video fingerprint are disclosed. The method includes: performing shot boundary detection on content of a video; determining a time duration of each shot according to positional points of the shot boundary, and compose the time duration of each shot into a shot boundary time slice sequence; and obtaining video fingerprint information according to the time slice sequence.
Claims14
A method implemented by one or more computing devices, the method comprising:
- performing shot boundary detection on content of a video;
- determining a time duration of each shot according to positional points of a shot boundary, and compose the time duration of each shot into a shot boundary time slice sequence;
- and
- obtaining video fingerprint information according to the time slice sequence, wherein obtaining the video fingerprint information according to the time slice sequence comprises:
- obtaining video fingerprint information of a current shot according to a difference between a time slice element of the current shot and a time slice element of an adjacent shot in the time slice sequence, obtaining the video fingerprint information of the current shot comprising:
- separately calculating quantized differences of the time slice element of the current shot with a time slice element of a previous shot and a time slice element of a next shot in the time slice sequence;
- using the quantized differences corresponding to the time slice element of the current shot as an index item of the time slice element of the current shot;
- and
- using the index item of the time slice element of the current shot, a video sequence number of the video and a shot sequence number of a time slice of the current shot as the video fingerprint information of the current shot.
An apparatus comprising:
- one or more processors;
- memory storing executable instructions that, when executed by the one or more processors, cause the one or more processors to perform acts comprising:
- performing shot boundary detection on content of a video, wherein performing the shot boundary detection on the content of the video comprises:
- calculating a normalized histogram of each image frame;
- calculating a normalized cumulative histogram of each image frame;
- and
- calculating a difference between cumulative histograms of every two adjacent image frames to form a difference sequence;
- smoothing the difference sequence using a window of a preset size;
- calculating a standard deviation of the difference sequence within a time window whose time length is a preset length;
- determining a time duration of each shot according to positional points of a shot boundary, and composing the time duration of each shot into a shot boundary time slice sequence;
- and
- obtaining video fingerprint information according to the time slice sequence.
One or more computer readable media storing executable instructions that, when executed by one or more processors, cause the one or more processors to perform acts comprising:
- performing shot boundary detection on content of a video;
- determining a time duration of each shot according to positional points of a shot boundary, and compose the time duration of each shot into a shot boundary time slice sequence;
- and
- obtaining video fingerprint information according to the time slice sequence, wherein obtaining the video fingerprint information according to the time slice sequence comprises:
- obtaining video fingerprint information of a current shot according to a difference between a time slice element of the current shot and a time slice element of an adjacent shot in the time slice sequence, wherein obtaining the video fingerprint information of the current shot comprises:
- separately calculating quantized differences of the time slice element of the current shot with a time slice element of a previous shot and a time slice element of a next shot in the time slice sequence;
- using the quantized differences corresponding to the time slice element of the current shot as an index item of the time slice element of the current shot;
- and
- using the index item of the time slice element of the current shot, a video sequence number of the video and a shot sequence number of a time slice of the current shot as the video fingerprint information of the current shot.
Dependencies are inferred from the claim text.