US 2021/0281891 A1 · PARTIAL-VIDEO NEAR-DUPLICATE DETECTION

Abstract

Methods, systems, and computer programs are presented for detecting near duplicates and partial matches of videos. One method includes an operation for receiving a video containing frames. For each frame, keypoints are determined within the frame. For each keypoint, a horizontal gradient vector is calculated based on a horizontal gradient at the keypoint and a vertical gradient vector is calculated based on a vertical gradient at the keypoint. The horizontal and vertical gradients are binary vectors. Further, a keypoint description is generated for each keypoint based on the horizontal gradient vector and the vertical gradient vector. Further, the frames are matched to frames of videos in a video library based on the keypoint descriptions of the keypoints in the frame in the videos in the video library. Further, a determination is made if the video has near duplicates in the video library based on the matching.

Claims20

Dependent claims

A method comprising:

receiving, by one or more processors, a video containing a plurality of frames;
for each frame, determining keypoints within the frame;
for each keypoint, calculating, by the one or more processors, a horizontal gradient vector based on a horizontal gradient at the keypoint and a vertical gradient vector based on a vertical gradient at the keypoint, the horizontal gradient vector and the vertical gradient vector being binary vectors;
for each keypoint, generating, by the one or more processors, a keypoint description based on the horizontal gradient vector and the vertical gradient vector;
for each frame, matching, by the one or more processors, the frame to frames of videos in a video library based on the keypoint descriptions of the keypoints in the frame and keypoint descriptions of the keypoints in frames of the videos in the video library;
and
determining, by the one or more processors, if the video has near duplicates in the video library based on the matching.

4 more elements

A system comprising:

a memory comprising instructions;
and
one or more computer processors, wherein the instructions, when executed by the one or more computer processors, cause the system to perform operations comprising:
receiving a video containing a plurality of frames;
for each frame, determining keypoints within the frame;
for each keypoint, calculating a horizontal gradient vector based on a horizontal gradient at the keypoint and a vertical gradient vector based on a vertical gradient at the keypoint, the horizontal gradient vector and the vertical gradient vector being binary vectors;
for each keypoint, generating a keypoint description based on the horizontal gradient vector and the vertical gradient vector;
for each frame, matching the frame to frames of videos in a video library based on the keypoint descriptions of the keypoints in the frame and keypoint descriptions of the keypoints in frames of the videos in the video library;
and
determining if the video has near duplicates in the video library based on the matching.

7 more elements

A non-transitory machine-readable storage medium including instructions that, when executed by a machine, cause the machine to perform operations comprising:

receiving a video containing a plurality of frames;
for each frame, determining keypoints within the frame;
for each keypoint, calculating a horizontal gradient vector based on a horizontal gradient at the keypoint and a vertical gradient vector based on a vertical gradient at the keypoint, the horizontal gradient vector and the vertical gradient vector being binary vectors;
for each keypoint, generating a keypoint description based on the horizontal gradient vector and the vertical gradient vector;
for each frame, matching the frame to frames of videos in a video library based on the keypoint descriptions of the keypoints in the frame and keypoint descriptions of the keypoints in frames of the videos in the video library;
and
determining if the video has near duplicates in the video library based on the matching.

4 more elements

Dependencies are inferred from the claim text.

Partial-video near-duplicate detection

Life of this patent

Partial-video near-duplicate detection

Life of this patent

Abstract

Claims20