20240054786. VIDEO STREAM MANIPULATION simplified abstract (Neatframe Limited)
Contents
VIDEO STREAM MANIPULATION
Organization Name
Inventor(s)
Simen Andresen of Fornebu (NO)
Håkon Skramstad of Fornebu (NO)
VIDEO STREAM MANIPULATION - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240054786 titled 'VIDEO STREAM MANIPULATION
Simplified Explanation
The method described in the patent application involves manipulating a video stream captured by a camera in a video conferencing endpoint into multiple views corresponding to regions of interest. This is achieved by detecting objects of predefined types in a frame of the initial video stream, selecting crop regions from the frame, and transmitting these crop regions, each containing at least one bounding box with a detected object of a predefined type.
- Detect objects of predefined types in a frame of the initial video stream
- Select crop regions from the frame containing the detected objects
- Transmit the crop regions, each with at least one bounding box
Potential Applications
- Video conferencing systems
- Security surveillance systems
- Virtual reality applications
Problems Solved
- Enhances user experience in video conferencing by providing multiple views of regions of interest
- Improves security surveillance by focusing on specific objects of interest
- Enhances immersion in virtual reality applications by dynamically adjusting views based on detected objects
Benefits
- Improved user experience in video conferencing
- Enhanced security surveillance capabilities
- Increased immersion and realism in virtual reality applications
Original Abstract Submitted
a method for manipulating an initial video stream captured by a camera in a video conferencing endpoint into multiple views corresponding to regions of interest. the method comprises detecting objects of one or more predefined types in a frame of the initial video stream; selecting a plurality of crop regions () from the frame of the initial video stream, and transmitting the plurality of crop regions. each crop region includes at least one bounding box and each bounding box includes a detected object of a predefined type.
