Intelligent video automatic editing method

Abstract

The invention discloses an intelligent video automatic editing method, which comprises the following steps of: 1, performing semantic analysis on a target text, and selecting and matching corresponding video clips in a video library; 2, performing correlation and similarity analysis on the video clips, and performing detail comparison at the same time; 3, previewing and displaying a video clip combination finished product, and manually modifying or adjusting the same; and 4, seamlessly splicing and rendering the plurality of video clips, and exporting a required video. Compared with the priorart, the method has the following beneficial effects that the manual link is reduced, and the production cost is reduced; the person participates in assisting algorithm learning instead of participating in editing originally; the algorithm retrieval time is shorter, the recommendation algorithm is accurate, and the recall is higher than that of the human brain; and in the aspect of final quality,the consistency of the video is better, and instability caused by human factors is avoided.

Description

technical field [0001] The invention relates to the technical field of video editing, in particular to a method for automatic mixing and cutting of intelligent video. Background technique [0002] Video mixing is a process of dividing single or multiple videos, extracting target segments, and then recombining them to generate new videos; the existing video mixing methods are mainly manual editing and machine-assisted editing. [0003] Manual editing is a traditional method. Humans edit one or more target videos and synthesize the required video clips. The editing, adding text material, adding animation material and video synthesis must use one or more specific software. Machine-assisted editing, which is more intelligent than manual editing, is that the machine parses one or more target videos provided by humans, separates the audio track and the picture, converts the audio track into text, and then analyzes the keywords in the text. , recommend relevant video clips based o...

AI Technical Summary

Problems solved by technology

[0003] Manual editing is the traditional way, one or more target videos are edited by humans, and the required video clips are synthesized together. One or more specific software is used for editing, adding text materials, adding animation materials and video synthesis to complete; machine-assisted editing is smarter than manual editing. It is a machine that analyzes one or more target videos provided by people, separates the audio track and picture, converts the voice track into text, and then analyzes the keywords in the text. , recommend relevant video clips based on keywords, and finally cut and edit the video by humans. When compositing, the machine will render the video according to the established template, such as adding background music, adding a fixed title and ending; but the above method, The efficiency of retrieving video clips is low, the recommendation method is rough when matching keywords, and it is difficult to define how long the clips are, resulting in rough granularity of video clips, and the quality of video editing depends entirely on the editors, resulting in unstable quality. Mixed video clipping; the current technical solution is to input text or article links and automatically convert them into videos, but the main application field of this solution is entertainment short videos, and the correlation and similarity between articles and videos are relatively high. Low, the content of the video and the text may be inconsistent; in the professional field, there are high requirements for the professional authority of the segment and the text, and the text needs to have a high correlation with the target segment, and the text corresponds to the target segment one-to-one

Images