Fused Background Reconstruction for Halo-Free Video Conferencing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video conferencing systems suffer from a halo effect around the foreground subject due to insufficient background blurring, which is caused by a lack of knowledge about obstructed background visual information and resulting matte generation.

Innovation Solution

The system determines a fused background by combining multiple background images over time, refining it to reconstruct the background in the absence of the foreground, thereby mitigating the halo effect by increasing background visibility and reducing the appearance of halos.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If simple background blurring is applied, then processing speed is maintained, but halo effect appears around the foreground subject

Engineering Contradiction:
Improveprocessing speedVSAvoidhalo effect
Core Design Contradiction:
ProductivityVSObject-affected harmful factors

Solution Approach 1:

The system performs preliminary actions by capturing multiple background frames before the foreground subject moves, then reconstructs the complete background in advance. This pre-reconstruction of the background using multiple frames eliminates the need for complex real-time processing during video conferencing, maintaining processing speed while preventing the halo effect through accurate background representation.

Inventive Principle:
Principle #10Preliminary action

2Loss of information

If multiple background frames are combined to reconstruct the complete background, then background visibility is improved, but processing complexity increases

Engineering Contradiction:
Improvebackground visibilityVSAvoidprocessing complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The system segments the image into foreground and background components, then processes only the background portion by combining multiple background frames. This segmentation approach allows the system to reconstruct the complete background using multiple frames to improve visibility, while limiting complex processing to only the background region, thereby managing overall processing complexity efficiently.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS12536617B2Fused images backgrounds
Publication Date: 2026.01.27 HEWLETT PACKARD DEVELOPMENT COMPANY LP
  • US12536617B2 patent drawing
  • US12536617B2 patent drawing
  • US12536617B2 patent drawing

AI summary

In some examples in accordance with the present description, an electronic device is provided. The electronic device includes an image sensor and a controller. The controller is to receive multiple images from the image sensor including respective foregrounds and backgrounds. The controller is also to segment the respective foregrounds from the backgrounds. The controller is also to combine the backgrounds to form a fused background. The controller is also to receive a new image including a new image foreground and a new image background. The controller is also to segment the new image foreground from the new image background.