Video Close-Up Box Merging to Prevent Duplicate Participant Views

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In video communication, participants who are close to each other often appear in multiple close-up boxes, affecting the viewing experience and communication efficiency due to repeated appearances.

Innovation Solution

A method for determining close-up pictures involves recognizing close-up boxes for head objects, selecting and updating these boxes based on principal objects, and ensuring that each participant appears in only one box by deleting redundant boxes, while maintaining a rational layout.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If close-up boxes are constructed for each participant individually, then each participant can be clearly seen in their own box, but participants who are close to each other will appear in multiple boxes causing repetition

Engineering Contradiction:
Improveparticipant visibilityVSAvoidrepeated appearance
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent merges close-up boxes by detecting when a second head object appears in a close-up box constructed for another participant. When detected, the second head object's close-up box is merged into the first participant's close-up box, preventing the second participant from appearing in separate repeated boxes while maintaining clear visibility of both participants in the merged box.

Inventive Principle:
Principle #5Merging (Combining)

2Ease of operation

If close-up boxes are constructed for each participant individually, then each participant has their own dedicated box, but the device complexity increases due to managing multiple boxes

Engineering Contradiction:
Improveindividual close-up constructionVSAvoidbox management
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The patent implements dynamic close-up box management where the system initially constructs individual close-up boxes for each participant, then dynamically merges boxes when head objects are detected in multiple boxes. This dynamic approach maintains the simplicity of individual box construction while reducing the final number of boxes to display, optimizing both ease of operation and device complexity.

Inventive Principle:
Principle #15Dynamics

3Quantity of substance

If all close-up boxes are displayed, then complete information about all participants is shown, but the viewing experience deteriorates due to repeated appearances

Engineering Contradiction:
Improveinformation completenessVSAvoidviewing experience
Core Design Contradiction:
Quantity of substanceVSObject-affected harmful factors

Solution Approach 1:

The patent extracts and removes redundant close-up boxes from the display set. When a second head object is detected in a close-up box, the system extracts the second participant's information into the first participant's close-up box and removes the second participant's separate close-up box from the display, eliminating repetition while preserving complete participant information.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP4601285A1Close-up picture determination method and apparatus, device, and storage medium
Publication Date: 2025.08.13 GUANGZHOU SHIYUAN ELECTRONICS CO LTD
  • EP4601285A1 patent drawingFigure 1~4
  • EP4601285A1 patent drawingFigure 5~7
  • EP4601285A1 patent drawingFigure 8~9

AI summary

Embodiments of the present disclosure disclose a method apparatus, device for determining a close-up picture and storage medium. The method includes: recognizing a close-up box to which each head object in the current frame of video stream data belongs, where each close-up box corresponds to at least one principal object, and the principal object is the head object that has been enclosed by the corresponding close-up box; selecting a close-up box from the close-up boxes as the current close-up box; determining that there is a second head object capable of being added to the current close-up box, determining the second head object as the principal object corresponding to the current close-up box, and deleting the close-up box constructed for the second head object; selecting a second close-up box to update the current close-up box as the selected second close-up box, and returning to perform an operation of when there is a second head object capable of being added to the current close-up box, until all current close-up boxes are traversed; and displaying the close-up picture of each close-up box in the video stream data. This method can solve a technical problem of repeatedly appearing close participants in multiple close-up boxes during close-up construction of video communication participants.