A Progress Recognition Method Based on Multiple Video Sources

Abstract

The invention discloses a progress recognition method based on a plurality of video sources. The progress recognition method comprises the following steps: 1, building a video resource library, preprocessing video resources, and building a feature code time pair database; 2, randomly exporting a video A and a feature code time pair corresponding to the video A, and matching all feature codes of the video A with all feature codes in the feature code time pair database to obtain another video B which has most feature codes being the same as the feature codes of the video A; 3, acquiring P feature codes occurring in both of the video A and the video B; and 4, building a rectangular coordinate system by taking a time dimension of the video B as a reference, and performing fitting processing by taking the time dimension of the video B as an axis x and taking a time dimension of the video A as an axis y, wherein an intercept parameter obtained by fitting is a time progress difference value between the video A and the video B. Through adoption of the progress recognition method, the calculation amount during duplication removal of different video sources is small, and the technical effect of recognizing videos of a plurality of versions can be achieved.

Description

technical field [0001] The invention relates to the field of video information processing, in particular to a method for identifying progress based on multiple video sources. Background technique [0002] The same film and television drama, in addition to different versions such as the full version, the deleted version, and the skipped opening and ending version, there are also differences in format, and even the timing of commercial insertion in the same scene shot is also different. These film and television works or programs are played through various media channels, and different video sources form a huge-capacity multimedia database. [0003] Because the streaming media itself has the characteristics of fast transmission speed and real-time transmission, users don’t have to wait until the entire video file is downloaded before watching it. They only need to order it online on the streaming media and watch it after tens of seconds of delay. Similarly, The push and inter...

AI Technical Summary

Problems solved by technology

[0009] The present invention provides a progress recognition method based on multiple video sources, which solves the problem that in the prior art, the existing video processing methods cannot uniformly locate the progress of video sources and cannot effectively arrange different video sources. Heavy weight, and the technical problem of not being able to identify the progress of different video versions, it has realized the unified positioning of the progress of video sources, the calculation of different video sources is small, and the identification of multi-version videos is possible. technical effect

[0010] In order to solve the above-mentioned technical problems, the embodiment of the present application provides a progress recognition method based on multiple video sources. The present invention aims to solve the problem of large amount of calculation in multimedia video database video re-ranking, that is, when we perform matching, in Use a certain method in the established feature library to find video sources that may be related (first perform rough matching), and then perform accurate progress matching

Images