Composite Video Generation via Foreground Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional techniques for generating composite video frames are computationally inefficient and produce undesirable visual artifacts.

Innovation Solution

The method involves foreground-background segmentation using a predictive model to extract and transform foreground object images based on camera motion, applying partial transparency and transformations such as rotation, translation, and scaling to generate composite images or videos from a single video or image sequence, preserving background motion and improving computational efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional techniques are used to generate composite video frames, then composite images can be produced, but computational efficiency deteriorates and visual artifacts increase

Engineering Contradiction:
Improvecomputational efficiencyVSAvoidvisual artifacts
Core Design Contradiction:
ProductivityVSObject-generated harmful factors

Solution Approach 1:

The patent applies segmentation by dividing the image processing into foreground object extraction and background separation. Foreground objects are identified and extracted from video frames, then transformed based on camera motion while the background remains static. This segmentation approach reduces computational burden by processing only relevant foreground elements and eliminates visual artifacts by properly separating moving objects from the background scene.

Inventive Principle:
Principle #1Segmentation

2Manufacturing precision

If foreground object images are transformed based on camera motion, then object path depiction improves, but computational complexity increases

Engineering Contradiction:
Improveobject path accuracyVSAvoidtransformation complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by pre-calculating camera motion parameters (translation, rotation, scaling) from the video sequence before transforming foreground objects. Motion estimation is performed on the background or reference frames first, and these pre-computed transformation parameters are then applied to foreground objects. This approach ensures accurate object path depiction while reducing computational complexity during the composite generation phase.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS11637971B2Automatic composition of composite images or videos from frames captured with moving camera
Publication Date: 2023.04.25 GOPRO INC
  • US11637971B2 patent drawing
  • US11637971B2 patent drawing
  • US11637971B2 patent drawing

AI summary

A processing device generates composite images from a sequence of images. The composite images may be used as frames of video. A foreground/background segmentation is performed at selected frames to extract a plurality of foreground object images depicting a foreground object at different locations as it moves across a scene. The foreground object images are stored to a foreground object list. The foreground object images in the foreground object list are overlaid onto subsequent video frames that follow the respective frames from which they were extracted, thereby generating a composite video.