Portrait Edge Fluctuation Offset for Mobile Video Special Effects

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video processing technologies for mobile devices face challenges in adding special effects due to high calculation requirements, making them unsuitable for mobile phones where traditional noise perturbation-based rendering technologies are too resource-intensive.

Innovation Solution

A method that separates portrait foregrounds from backgrounds, determines fluctuation offsets based on playback time and a fluctuation parameter, and superimposes these offsets onto the video frames to create a special effect, utilizing the principle of wave propagation to reduce computational load.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If traditional noise perturbation-based rendering technologies are used to add special effects to videos, then the special effect quality can be improved, but the calculation amount becomes too high for mobile phones to handle

Engineering Contradiction:
Improvespecial effect qualityVSAvoidcalculation amount
Core Design Contradiction:
Manufacturing precisionVSUse of energy by moving object

Solution Approach 1:

The patent extracts only the essential components needed for special effect generation (portrait edge information) rather than applying full noise perturbation processing. By separating and processing only the edge portions of portrait foregrounds, the calculation amount is dramatically reduced while maintaining visual quality, making it suitable for mobile phone processing capabilities

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments the video processing into distinct steps: separating portrait foregrounds from backgrounds, extracting edge information from these foregrounds, and then applying fluctuation offsets only to the extracted edge portions. This segmentation allows complex special effects to be achieved through multiple simple processing stages rather than one computationally intensive operation

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS11410342B2Method for adding special effect to video, electronic device and storage medium
Publication Date: 2022.08.09 BEIJING DAJIA INTERNET INFORMATION TECH CO LTD
  • US11410342B2 patent drawing
  • US11410342B2 patent drawing
  • US11410342B2 patent drawing

AI summary

Embodiments of the present disclosure provide a method for adding a special effect to a video, an electronic device and a storage medium. First portrait edge information of a first frame is determined based on the portrait foregrounds of the first frame and a previous frame. A fluctuation offset based on a playback time and a fluctuation parameter is added to the first portrait edge information to obtain second portrait edge information. The second portrait edge information is superimposed on the first frame to obtain a second frame having the special effect.