Video Style Migration for Short-Duration Time-Lapse Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing time-lapse photography methods require users to fix their electronic devices in one place for extended periods, limiting the scene, device, and duration of shooting, making it difficult to capture rapid changes like day-to-night transitions in a short time.
Innovation Solution
An electronic device performs frame extraction and style migration on a video, using fused style migration models to create a time-lapse effect, allowing users to capture long-time video shooting without fixation, and includes image stabilization to enhance user convenience and interest.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional time-lapse photography is used to capture long-time scene changes, then the time-lapse effect is achieved, but the user must fix the electronic device in one place for extended periods, limiting scene adaptability and shooting duration
Solution Approach 1:
The patent segments the time-lapse photography process into two distinct phases: (1) rapid sequential shooting of multiple images in a short duration, and (2) post-processing through frame extraction and style migration to simulate long-time effects. This segmentation allows the shooting phase to be completed quickly without requiring prolonged device fixation, thereby improving scene adaptability while maintaining the time-lapse effect.
Solution Approach 2:
The patent applies style migration as a preliminary processing step that transforms rapidly captured images into styles resembling long-exposure time-lapse effects. By pre-defining multiple styles (e.g., daytime, nighttime, different weather conditions) and using AI models to migrate between them, the system creates the appearance of extended shooting duration without actually requiring the device to remain fixed for long periods, thus resolving the contradiction between shooting duration and scene adaptability.
2Reliability
If frame extraction processing is performed on video to achieve time-lapse effect, then the time-lapse video is produced, but the user needs to fix the electronic device for long time, imposing high limitation on scene and device
Solution Approach 1:
The patent replaces the mechanical requirement of physical device fixation with an AI-based style migration system. Instead of requiring the device to remain stationary for extended periods to capture authentic time-lapse sequences, the system uses neural network models to transform rapidly captured images into styles that simulate long-duration time-lapse effects. This substitution eliminates the need for mechanical stability during extended shooting while maintaining the visual quality and reliability of time-lapse effects.
Solution Approach 2:
The patent changes the parameter of shooting duration from extended periods to short durations, while compensating for the time-lapse effect through post-processing style migration. By adjusting the style parameters (e.g., lighting conditions, atmospheric effects) through AI transformation rather than capturing them over actual extended time periods, the system maintains time-lapse effect quality while significantly reducing the ease of operation constraints related to device fixation.
3Productivity
If style migration processing is applied to first image sequence, then time-lapse effect is achieved in short time, but multiple style migration models need to be fused, increasing processing complexity
Solution Approach 1:
The patent merges multiple style migration models into a unified processing framework that can handle sequential style transformations. By combining multiple pre-trained style migration models (each specialized in transforming to a specific style such as daytime, nighttime, or different weather conditions) into an integrated system, the patent enables efficient style migration across multiple frames while managing the complexity through systematic model fusion strategies.
Solution Approach 2:
The patent implements dynamic style migration where the style transformation parameters can be adjusted and optimized during processing. The system dynamically selects and applies appropriate style migration models based on the input image characteristics and desired output style, allowing flexible adaptation without requiring all models to be simultaneously active, thereby improving shooting efficiency while managing processing complexity through dynamic model selection rather than static full-model deployment.
Data Source
AI summary
An image processing method includes an electronic device configured to perform style migration processing on a first image sequence based on a target migration style by using a fused style migration model into which a plurality of single-style migration models is fused in order to obtain a second image sequence. A style of a 1st frame of image to a style of a last frame of image in the second image sequence change in a first style order in styles of output images of the plurality of single-style migration models. The first image sequence may be from a video shot by using the electronic device. The electronic device may save a plurality of frames of images in the second image sequence as a video. The video may present an effect of rapid time lapse during play.