Motion-Feature Video Synchronization Without Camera Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video synchronization methods require manual techniques or synchronized cameras, introducing unnecessary elements and are not event-specific, necessitating a solution for automated synchronization based on video subject movements.
Innovation Solution
An automated system using machine learning models to synchronize videos by analyzing 2D and 3D motion features of subjects, without requiring camera calibration or synchronization, enabling synchronization of videos from different perspectives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If manual synchronization techniques are used, then video synchronization can be achieved, but the process requires unnecessary human intervention and is not event-specific
Solution Approach 1:
The system extracts motion features directly from the video content itself to perform synchronization, allowing the videos to synchronize themselves based on their inherent motion characteristics without external intervention or pre-configured event parameters
Solution Approach 2:
The patent replaces manual mechanical synchronization operations with automated computer-based motion feature extraction and comparison, substituting human operators with algorithmic processing that automatically identifies corresponding motion patterns across videos
2Reliability
If synchronized cameras are used, then video synchronization can be achieved, but expensive specialized equipment is required
Solution Approach 1:
The patent extracts synchronization information directly from the video content by identifying and comparing motion features of subjects, separating the synchronization function from the camera hardware and making it independent of specialized synchronized camera equipment
Solution Approach 2:
The system creates a digital representation of motion features from video frames and uses these copied motion patterns to establish synchronization, replacing the need for physical camera synchronization mechanisms with digital feature matching
3Productivity
If conventional synchronization features are used, then video merging can be performed, but extraneous elements not related to the event are introduced
Solution Approach 1:
The patent focuses synchronization on local motion features of specific subjects within the videos rather than using global synchronization signals, allowing the system to identify and synchronize only the relevant event-specific motion patterns while ignoring extraneous background elements
Solution Approach 2:
Instead of using traditional approaches that start with synchronization signals or metadata and then match videos, the patent inverts the approach by extracting motion features from video content and using those features to drive the synchronization process, eliminating reliance on extraneous synchronization elements
Data Source
AI summary
A system includes a hardware processor and a system memory storing software code. The hardware processor is configured to execute the software code to receive a plurality of video sequences each generated by a respective one of a plurality of cameras, each of the video sequences depicting a respective one of a plurality of three-dimensional (3D) movements of a subject from a different perspective, extract, from each of the video sequences, a respective set of two-dimensional (2D) motion features, to provide a plurality of 2D motion feature sets, and synchronize, using the 2D motion feature sets, the plurality of video sequences.