Video abstract generating method and device
A video summary and frequency technology, applied in the field of image processing, can solve the problems of irrelevant targets interfering with video browsing, interference, and distracting the attention of public security staff
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0065] Embodiment 1 of the present invention provides a video summary generation method, see figure 1 shown, including steps:
[0066] Step S110, perform foreground detection on each frame of image, and detect objects in the foreground.
[0067] Foreground detection can use various related algorithms, such as mixed Gaussian background model, SACON (SAMPLE CONSENSUS), etc., which are not listed in this embodiment.
[0068] Step S111 , tracking the detected target, and extracting the motion track of the target.
[0069] In the tracking process, various algorithms can also be used, such as relatively simple nearest neighbor method, multi-target tracking algorithm, edge tracking algorithm and so on.
[0070] Step S112, dividing the object into different categories according to the color information of the object.
[0071] Since different target objects often have different color characteristics, different targets can be distinguished according to the color information of each t...
Embodiment 2
[0077] Embodiment 2 of the present invention provides a video abstract generation method, see figure 2 shown, including steps:
[0078] Step S210: performing background modeling on the image of the intended area by using a mixture of Gaussian functions, and extracting objects in the foreground.
[0079] Use mixed Gaussian to model the background of the image and extract the moving foreground. The number of mixed Gaussian functions can be selected according to the video scene, and a Gaussian model can be trained separately for shadows.
[0080] The single Gaussian background modeling function is
[0081] f ( x ; μ ; σ ) = φexp ( - ( x - μ ) ...
Embodiment 3
[0150] Embodiment 3 of the present invention provides a device for generating a video summary, see image 3 , including a foreground detection module, a tracking module, a classification module and a generation module.
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com