A method for extracting video key frames and close-up segments

Abstract

The present invention provides a method for extracting video key frames and close-up segments, comprising S1, obtaining video frames from a video source; S2, inputting video frames into a buffer area for buffering; S3, converting the first video frame and the last video frame in the buffer area A video frame is grayscaled respectively; S4, the edge area acquisition and the main body area acquisition are respectively performed on the first grayscale image and the second grayscale image; S5, the first similarity is calculated, and the second similarity is calculated; S6 According to the first similarity and the second similarity, it is judged whether all the video frames buffered in the buffer area form a close-up segment; S7, it is judged whether the close-up segment is a large-area blank video segment; S8, the last video frame is used as a key frame , to output the video segment consisting of all the video frames in the buffer as a close-up segment. The present invention requires few computing resources, low algorithm development cost, and small calculation amount, and at the same time, the attribute of the video source is also taken into consideration.

Description

technical field [0001] The invention relates to the field of video processing, in particular to a method for extracting video key frames and close-up segments. Background technique [0002] In the field of video user generated content (UGC-User Generated Content), valid video content is contained in a large number of invalid videos. User-generated content generally contains video images of different scenes than fixed-purpose video footage. The purpose of video key frame and segment extraction is to reduce the workload of image processing, the computing resources of image model recognition and shorten the output time of results. [0003] The current industry's method of extracting key frames and scenes in video: 1) only for specific types of objects and scenes. For example, crowds, iron towers. At this time, the key frame can be obtained by directly calling the AI ​​classification model; 2) The key frame is judged by the image as a whole, and whether the image belongs to t...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to disclose a method for extracting video key frames and close-up segments, which solves the problem of unnecessary computing resource consumption and extraction algorithm development costs caused by AI model calculations in the current video key frame extraction, based on the similarity of the overall image Analyzing technical issues that require a large amount of computation and do not take into account the shooting conditions

Images