Video Crosstalk Reduction via Temporal Dimming Sequences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video conferencing systems face the issue of video crosstalk, where the camera captures images of the local participant through the screen along with the projected images of the remote participant, leading to corrupted video streams and impaired eye contact and gaze awareness.
Innovation Solution
The method involves projecting the remote participant's video stream onto the screen with a predetermined dimming sequence, using two successive dimming factors K1 and K2 to adjust the intensity of the images, allowing for the reduction of video crosstalk by processing pairs of images to generate a crosstalk-free video stream.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the camera captures images through the screen to enable natural interaction and eye contact, then the interaction naturalness is improved, but video crosstalk occurs causing image corruption
Solution Approach 1:
The patent applies periodic action by using a temporal dimming sequence where the projected image intensity is varied over time (bright/dim/bright phases). The camera captures multiple frames at different dimming levels, and through temporal processing, the system separates the reflected participant image from the projected image, eliminating crosstalk while maintaining the screen-based interaction setup.
Solution Approach 2:
The patent uses preliminary action by pre-calculating and applying a dimming sequence to the projected image before capture. The system knows in advance the intensity pattern that will be applied, allowing it to prepare the corresponding unmixing weights and processing parameters needed to separate the images during post-processing, thus preventing crosstalk artifacts.
2Illumination intensity
If the remote participant's image is projected onto the screen with high intensity, then the visibility is improved, but the video crosstalk corruption increases
Solution Approach 1:
The patent applies dynamics by making the projected image intensity time-varying through a dimming sequence rather than maintaining constant high intensity. The system dynamically adjusts brightness levels (bright/dim/bright phases) and processes the temporal variations to separate the projected image from the reflected participant image, thus maintaining visibility while reducing crosstalk corruption.
Solution Approach 2:
The patent uses parameter changes by modifying the intensity parameter of the projected image over time according to a predetermined dimming sequence. By changing the brightness parameter dynamically and knowing the exact sequence of changes, the system can mathematically separate the projected content from the reflected participant image during post-processing.
3Manufacturing precision
If multiple frames are captured and processed to reduce crosstalk, then the video quality is improved, but the processing complexity increases
Solution Approach 1:
The patent applies feedback by using the known dimming sequence as a reference signal in the post-processing stage. The system captures multiple frames corresponding to different dimming levels and uses the predetermined intensity pattern as feedback information to calculate unmixing weights that separate the reflected and projected images, improving video quality through controlled temporal variations.
Solution Approach 2:
The patent uses preliminary action by pre-defining the dimming sequence and calculating the corresponding processing parameters before image capture. This allows the system to prepare the unmixing weights and separation algorithms in advance, reducing the computational complexity during real-time processing while still achieving high video quality through multi-frame analysis.
Data Source
AI summary
Methods and systems that reduce video crosstalk in video streams sent between participants in a video conference are disclosed. In one aspect, a method for reducing video crosstalk in a video stream sent from a local site to a remote site includes projecting a video stream of the remote site onto a screen at the local site. Each image in the video stream is dimmed according to a dimming factor of a dimming sequence. Crosstalk images of the local site are captured through the screen. Each crosstalk image is a blending of the image of the local site captured through the screen with a dimmed image of the remote site projected onto the screen. Images of the local site with reduced crosstalk are computed based on the dimming sequence. A video stream composed of the images of the local site with reduced crosstalk are sent to the remote site.