Video Conference Endpoint Spatial Boundary Auto-Framing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video conferencing systems struggle to reliably detect and frame individuals within a defined spatial boundary, particularly in environments with glass walls or open spaces, leading to unwanted participants being included in the video call.
Innovation Solution
A method and endpoint that define a spatial boundary using distance and angular criteria, identify individuals within the field of view, estimate their positions, and generate cropped video signals for transmission, excluding those outside the boundary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If auto-framing algorithms detect and frame all people in the field of view, then the completeness of capturing participants is improved, but the reliability of distinguishing intended participants from unwanted people is worsened
Solution Approach 1:
The patent segments the field of view into multiple spatial zones by defining a spatial boundary (e.g., virtual wall, plane, or volume) that divides the scene into an included region and an excluded region. This segmentation allows the system to selectively frame only people within the boundary while excluding others, resolving the contradiction between completeness and reliability.
Solution Approach 2:
The patent introduces a spatial boundary as an intermediary element between the camera and the people in the scene. This virtual boundary acts as a mediator that filters which people are framed, enabling the system to distinguish intended participants from unwanted people without losing information about those outside the boundary.
2Measurement precision
If the spatial boundary is defined using multiple parameters (distance, angle, volume), then the precision of participant identification is improved, but the device complexity is worsened
Solution Approach 1:
The patent implements dynamic spatial boundaries that can be adjusted in real-time based on detected people's positions and behaviors. The boundary parameters (distance, angle, volume) are not fixed but can be modified during the video conference, allowing the system to adapt to different meeting scenarios while maintaining precision without permanent complexity.
Solution Approach 2:
The patent changes the parameters defining the spatial boundary based on detected conditions. For example, the distance parameter can be adjusted based on how far people are from the camera, the angular extent can change based on their position in the field of view, and the boundary shape can transition between planar and volumetric forms to optimize participant identification precision.
Data Source
AI summary
A computer-implemented method of operating a video conference endpoint. The video conference endpoint includes a video camera which captures images showing a field of view. The method comprises: receiving data defining of a spatial boundary within the field of view, the spatial boundary being at least in part defined by a distance from the video camera; capturing an image of the field of view; identifying one or more persons within the field of view of the video camera; estimating a position of the or each person within the field of view of the video camera; and generating one or more video signals, which include one or more cropped regions corresponding to one or more persons determined to be within the spatial boundary, for transmission to a receiver.