Adaptive Video Image Allocation for Multi-Participant Conferencing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video conferencing technologies are limited by the fixed field of view of the camera, failing to adaptively adjust the real-time image display based on the presence and interaction of multiple participants in a meeting.
Innovation Solution
A method and apparatus that utilize object detection and sound localization to dynamically select and combine video images of target objects, generating a final image for display that prioritizes active speakers and adjusts based on spatial relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a fixed field of view camera is used for video conferencing, then the device structure is simple, but the adaptability of image allocation is poor
Solution Approach 1:
The system dynamically adjusts image allocation based on real-time detection of target objects and sound sources. The display switches between different camera feeds and layouts according to detected speaker positions and movements, making the previously static display system adaptive to changing meeting conditions.
Solution Approach 2:
The patent introduces an intermediary processing system that receives inputs from multiple cameras and microphones, processes this data through object detection and sound localization algorithms, and generates optimized display arrangements. This intermediary layer enables adaptability without requiring complex direct control of each camera.
2Loss of information
If only a single real-time image is displayed, then the device complexity is low, but the information completeness is insufficient
Solution Approach 1:
The system segments the display into multiple regions showing different camera feeds simultaneously. Instead of displaying a single real-time image, the interface divides the screen to show multiple participants' video streams, allowing users to see more information without requiring a single complex all-encompassing camera.
Solution Approach 2:
The display system is designed to serve multiple functions: it can show individual speaker close-ups, group views, and alternative camera angles within the same interface. This multi-functional display approach provides comprehensive information while using standard display hardware rather than requiring specialized equipment.
3Quantity of substance
If the real-time image is limited by the field of view of the lens, then the device structure is simple, but the coverage of target objects is insufficient
Solution Approach 1:
The system merges feeds from multiple cameras with different fields of view into a unified display. By combining the coverage areas of multiple lenses through software composition, the system achieves broader target object coverage without requiring a single complex wide-angle or panoramic camera system.
Data Source
AI summary
An allocation method of video images and a computing apparatus are provided. A plurality of video images are obtained. The video images correspond to a plurality of target objects. A detection result corresponding to the target objects is obtained. At least one target image is selected from the video images according to the detection result. A final image is generated according to the target image. The final image is used to be presented on a user interface. In this way, the screen content can be adjusted according to the situation.