Video summary generation method, server and client-terminal

Abstract

The invention brings forward a video summary generation method, a server and a client-terminal. The method comprises the steps that a video source file is decoded by a video summary server. Sampling is performed on the obtained original video frames. An image statistic characteristic is extracted from each sampling frame. Inter-frame distance of all the statistic characteristics is calculated with respect to every two adjacent sampling frames. The weighted value of the inter-frame distance of all the statistic characteristics of the two sampling frames is calculated. The weighted statistic characteristic inter-frame distance set of the video source is sent to a video client terminal. The video change curve of the video source is obtained by the video client-terminal according to the weighted statistic characteristic inter-frame distance set. The basic hierarchical structure of the video source is confirmed according to the video change curve. The basic hierarchical structure of the video source is adjusted to be a summary hierarchical structure meeting the requirements of users according to the video summary hierarchical requirements inputted by the users. Paragraph division is performed on the video source according to the summary hierarchical structure, and paragraph information acts as the video summary of the video source to be provided to the users. Processing burden of the video client-terminal is reduced by the method.

Description

technical field [0001] The invention relates to the field of multimedia technologies, in particular to a method for generating a video summary, a video summary server, and a video client. Background technique [0002] With the development of multimedia technology and the popularization of video capture equipment, video files appear more and more widely in people's life and work. The urgent needs of many video users, but just reading the text and picture posters of the video is far from satisfying the requirements of users to fully understand the composition structure and main content of the video, so the video summary, a new video description method, came into being , by browsing the summary of the video, users can understand the paragraph structure of the video and watch the highlights in each paragraph. [0003] Existing video summary generation methods mainly consist of two parts: [0004] 1) Feature extraction: first sample the video, such as 5 or 10 frames, and calcul...

AI Technical Summary

Problems solved by technology

[0007] 1) The multi-feature extraction algorithm requires a large number of calculations for each sampling frame, resulting in a slow overall processing speed, especially with the emergence of a large number of high-definition and full high-definition videos, the problem of slow extraction speed has been further aggravated, which has seriously affected users experience

[0008] 2) The current video summary generation method can only provide a unified browsing method, and cannot be customized according to user requirements to meet the differentiated viewing needs of different users in different time periods, thereby affecting the user's browsing experience

[0009] 3) The current video summary provides users with a single service function and lack of interaction, which makes users lack a sense of participation and is not conducive to the cultivation of user loyalty

Images