A video summarization method based on multi-feature fusion
A multi-feature fusion and video summarization technology, applied in image communication, selective content distribution, electrical components, etc., can solve the problems of unfavorable user browsing and unfavorable video content grasping of video summarization, so as to improve experience, save time and energy Effect
Active Publication Date: 2020-07-31
NANJING UNIV OF SCI & TECH
View PDF6 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
On the basis of the research on massive video data, the analysis, processing and storage of video data are lacking, which leads to the blindness of users when looking for useful video data. In addition, most of the results of generating video summaries are not Great ideal, because the video summaries generated by many methods are static video summaries, which are not conducive to users' browsing, and are not conducive to users' grasp of video content
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0019] combine figure 1 , a video summarization method based on multi-feature fusion of global importance and local importance, including the following steps:
[0020] Step 1, get the video and use the video as input data;
[0021] Step 2, the input video data is processed to obtain the number of segmentation points and video clips one by one;
[0022] Step 3, extract the video frame and the video frame central block in each video segment;
[0023] Step 4, performing feature and image quality calculation on the extracted video frame and video frame center block respectively;
[0024] Step 5, calculate the global importance and local importance according to the obtained features;
[0025] Step 6, fusing the obtained global importance and local importance of each frame to obtain the final fusion importance;
[0026] Step 7, calculating the importance of each video segment according to the segmentation point;
[0027] Step 8, according to the importance of each obtained vide...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention provides a multi-feature fusion type video abstract generation method. The method comprises the following steps: obtaining a video, and using the video as input data; performing fragment segmentation on the input video data, recording segmentation points and the number of video fragments; extracting a video frame and a video frame central block in each video fragment; calculating features and image quality of the extracted video frames and the extracted video frame central blocks; calculating global importance and local importance according to the obtained features; and fusing the obtained global importance and the obtained local importance of each frame to obtain fusion importance; calculating the importance of each video fragment according to the segmentation points; selecting the video fragments according to the obtained importance of each video fragment and a set threshold, and selecting an optimized video fragment subset; and synthesizing a video abstract according to the selected video fragment subset.
Description
technical field [0001] The invention relates to a food analysis and image processing technology, in particular to a multi-feature fusion video summary generation method. Background technique [0002] The current rapid development of Internet technology and smart devices has made the ways for people to obtain and browse videos more and more diversified. At the same time, they are faced with more and more video data. Faced with such a large amount of video data, how to find The video data or visual information we need is a current research hotspot, and it is also the research content of video analysis technology. On the basis of the research on massive video data, the analysis, processing and storage of video data are lacking, which leads to the blindness of users when looking for useful video data. In addition, most of the results of generating video summaries are not It is a great ideal, because the video summaries generated by many methods are static video summaries, which...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): H04N21/8549H04N21/44
CPCH04N21/44008H04N21/44016H04N21/8549
Inventor 李泽超唐金辉胡铜铃
Owner NANJING UNIV OF SCI & TECH