Automatic extraction method of the structure of a video sequence
US6628710B1 · kind B1 · utility
Assignee
Inventors
Key dates
| Filing date | Jul 6, 2000 |
| Grant date | Sep 30, 2003 |
| Priority date | — |
| Expiry date | Apr 29, 2022 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG11B27/28
- WIPO fieldAudio-visual technology
- WIPO sectorElectrical engineering
Abstract
The invention relates to a method intended to automatically create a description of a video sequence—i.e. its table of contents—, by means of an analysis of the sequence. The main steps of the method are a shot detection, a sub-division of these shots into sub-entities called micro-segments, and the creation of the final hierarchical structure of the processed sequence. According to the invention, the shot detection step computes the mean displaced frame difference curve, detects the highest peaks of said curve, removes, by filtering, some negative or positive peaks, extracts markers, and propagates them on the curve.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.