Video Conference System Sound Source Tracking Sub-Image Analysis

Solution Overview

Problem

Conventional video conference devices face challenges in efficiently tracking sound sources and displaying corresponding close-up images with minimal computation, especially in scenarios with multiple conference members, leading to high processor resource usage.

Innovation Solution

A video conference system comprising an image detection device, a sound source detection device, and a processor that outputs a positioning signal when a sound source is detected, allowing the processor to determine if a real face image exists in a sub-image block of the conference image and display a close-up image accordingly, reducing computation by analyzing only specific sub-image blocks.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the processor performs image analysis on the entire captured conference image to determine the location of a close-up face, then the accuracy of sound source tracking is improved, but the computation time and processor resource consumption increase significantly

Engineering Contradiction:

Improvesound source tracking accuracyVSAvoidcomputation time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent divides the conference image into multiple sub-images based on the positioning signal from the sound source detection device. Instead of analyzing the entire image, the processor only analyzes the specific sub-image where the sound source is located, thereby reducing computation time while maintaining tracking accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by focusing the image analysis only on the specific region (sub-image) corresponding to the sound source location. This allows the system to concentrate computational resources on the most relevant area rather than processing the entire image uniformly.

Inventive Principle:

Principle #3Local quality

2Measurement precision

If the processor performs image analysis on the entire captured conference image to determine the location of a close-up face, then the accuracy of sound source tracking is improved, but the processor resource consumption increases significantly

Engineering Contradiction:

Improvesound source tracking accuracyVSAvoidprocessor resource consumption

Core Design Contradiction:

Measurement precisionVSPower

Solution Approach 1:

The patent divides the conference image into multiple sub-images based on the positioning signal from the sound source detection device. Instead of analyzing the entire image, the processor only analyzes the specific sub-image where the sound source is located, thereby reducing computation time while maintaining tracking accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by focusing the image analysis only on the specific region (sub-image) corresponding to the sound source location. This allows the system to concentrate computational resources on the most relevant area rather than processing the entire image uniformly.

Inventive Principle:

Principle #3Local quality

3Productivity

If the system automatically tracks sound sources and displays corresponding close-up images, then the video conference effectiveness is improved, but the device complexity increases

Engineering Contradiction:

Improvevideo conference effectivenessVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent integrates multiple functions into a unified system: the image detection device captures images, the sound source detection device identifies sound locations, and the processor coordinates both to automatically generate close-up conference images. This multi-functional integration improves video conference effectiveness while managing system complexity through coordinated operation of existing components.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS10931919B2Video conference system, video conference apparatus, and video conference method

Publication Date: 2021.02.23 CORETRONIC CORPORATION

AI summary

A video conference system, a video conference apparatus and a video conference method are provided. The video conference system includes a video conference apparatus and a display apparatus. The video conference apparatus includes an image detection device, a sound source detection device, and a processor. The image detection device obtains a conference image of a conference space. When the sound source detection device detects a sound generated by a sound source in the conference space, the sound source detection device outputs a positioning signal. The processor receives the positioning signal, and determines whether a real face image exists in a sub-image block of the conference image corresponding to the sound source according to the positioning signal to output the image signal. The display apparatus displays a close-up conference image including the real face image according to the image signal.