Method and device for extracting highlights of video

Abstract

The invention discloses a method and device for extracting highlights of a video, and the method comprises the steps: obtaining the number of bullet screens in a target video in each preset time period, adaptively determining a bullet screen number threshold value according to the number of bullet screens in the target video in each preset time period, and avoiding the subjectivity of threshold value selection; determining the preset time periods when the number of bullet screens in the target video is greater than the bullet screen number threshold value as highlight time periods; carrying out the clustering of all highlight time periods, and obtaining a plurality of clusters and the start and end moments of each cluster; carrying out the extraction of the target video according to the start and end moments of each cluster, and obtaining a target highlight segment. Through the clustering of the highlight time periods, the method achieves the discretization and fragmentization of highlight time points, guarantees that the target highlight segment has a certain time duration, and guarantees the wonderfulness of the target highlight segment.

Description

technical field [0001] The present invention relates to the technical field of video clipping, and more specifically, to a method and device for extracting video highlights. Background technique [0002] Watching videos has become an increasingly important part of the entertainment life of Chinese netizens, and online videos have increasingly become the most important way for people to watch videos. There are many online video resources, not only movies, TV, variety shows and other multimedia resources, but also micro movies, selfie videos and other resources. [0003] Videos watched by users are generally recommended by others or selected through popular web video rankings and hits to play. However, there are often some content in the video that is more attractive to users, and there are also some parts that are relatively boring and not of interest to users. For some of these boring parts, users will find that they are not interested after watching them, which not only af...

AI Technical Summary

Problems solved by technology

[0005] However, only finding out the time points when the number of bullet chats exceeds the threshold is prone to the problem of discrete and discontinuous time points. In this way, the time period taken out is very short, and there are few or no fragments with long time spans, and the effect is not good.

The problem with only finding out the period of time when the average number of barrages per second exceeds the threshold is that there may be a lot of barrages at a certain time point in the time interval, and the number of barrages at other time points is very small, so it is also possible that the entire interval The average number of barrages exceeds the threshold, but there is a large time difference between the time point with the most barrages and the time point with the lowest barrages in the entire interval, so the splendor of the continuous intervals taken out will be compromised

Images