Method and system for establishing ownership relation between sample video and copy video

Abstract

The invention relates to the computing field of video fingerprint, and discloses a method and a system for establishing an ownership relation between a sample video and a copy video. The method comprises the following steps: S1, adding the sample video to a sample database; S2, a video fingerprint similarity computing system synchronizing a similarity relation of the copy video relative to the sample video to a similarity relation database; S3, in accordance with an ownership strategy, computing the only ownership relation of the copy video; and S4, converting the only ownership relation of the copy video to an ownership relation between the sample video and the copy video. The invention also discloses a system for establishing an ownership relation between a sample video and a copy video. The method of the invention can conduct hierarchical computing on the ownership relation between the sample video and the copy video, and supports a plurality of sample databases and a plurality of fingerprint bottom layer service, converts the similarity relation of a video document to an ownership relation, and provides inquiry service to upper-layer application.

Description

technical field [0001] The present invention relates to the field of video fingerprint calculation, in particular to a method and system for establishing an attribution relationship between a sample video and a duplicate video. Background technique [0002] Video fingerprinting is a technology for software to identify, extract, and compress videos. It can generate a unique "fingerprint" to represent a video file. It is effectively used in audio data preprocessing, hash value comparison, and digital watermarking. On the basis of the existing video file fingerprint calculation service, the similarity relationship of video files is converted into the attribution relationship of sample video and copy video, which can realize the query of video-level attribution relationship for upper-layer applications, and provide the bottom layer for copyright and review services. support. [0003] The existing video library query system is based on a single underlying fingerprint service sys...

AI Technical Summary

Problems solved by technology

[0003] The existing video library query system is based on a single underlying fingerprint service system, which provides video-level attribution query for the upper-layer division system. However, the query system has the following problems: First, the query system only supports a single sample library. When multiple types of sample libraries are required, services cannot be provided; second, the query system can only access a single underlying fingerprint service system, and cannot summarize similar relationships when there are multiple sets of underlying fingerprint service systems; third, the There is no layering in the process of calculating the affiliation relationship in the query system, so if the underlying data is wrong or the system is abnormal, there will be data loss, and it will be very troublesome to rebuild the affiliation relationship; fourth, the attribution strategy of the sample library is fixed. When using the similarity relationship to calculate the attribution relationship, the attribution policy cannot be flexibly configured; fifth, when updating the video-level similarity relationship, it is not instantaneously switched, resulting in incomplete attribution relationship queried during the update period

[0004] With the stepwise increase in the amount of videos, the above-mentioned video library query system can no longer meet the management needs of existing videos, and it is urgent to develop a new system for establishing the ownership relationship between sample videos and copy videos

Images