Automatic video framing of conference participants
US10104338B2 · kind B2 · utility
Assignee
Inventor
Key dates
| Filing date | Aug 21, 2017 |
| Grant date | Oct 16, 2018 |
| Priority date | — |
| Expiry date | Aug 21, 2037 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH04N23/611
- WIPO fieldAudio-visual technology
- WIPO sectorElectrical engineering
Abstract
Systems and methods are described for automatically framing participants in a video conference using a single camera of a video conferencing system. A camera of a video conferencing system may capture video images of a conference room. A processor of the video conferencing system may identify a potential region of interest within a video image of the captured video images, the potential region of interest including an identified participant. Feature detection may be executed on the potential region of interest, and a region of interest may be computed based on the executed feature detection. The processor may then automatically frame the identified participant within the computed region of interest, the automatic framing including at least one of cropping the video image to match the computed region of interest and rescaling the video image to a desired resolution.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.