Full-Frame Media Stabilization With Motion-Adaptive Frame Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video stabilization methods in electronic devices suffer from suboptimal performance due to the use of fixed hyperparameters and camera motion agnostic parameters, leading to poorer stabilization, excessive crop, and loss of field of view, resulting in an unpleasant viewing experience.
Innovation Solution
A method for full frame media stabilization that dynamically selects stabilization procedures and frame fusion based on the type and variation of motion, using ego motion classification and adaptive path smoothing to preserve the field of view.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing video stabilization methods use fixed hyperparameters and camera motion agnostic parameters, then the stabilization process is simple, but the stabilization quality is poor and excessive crop occurs
Solution Approach 1:
The patent implements dynamic selection of stabilization procedures and frame fusion methods based on detected motion types and their variations. The system adapts the stabilization approach in real-time rather than using fixed parameters, allowing optimal stabilization quality for different motion scenarios while maintaining manageable complexity through automated adaptation.
Solution Approach 2:
The system changes parameters dynamically by selecting different stabilization procedures and frame fusion methods based on motion characteristics. Instead of using fixed hyperparameters, the system adjusts the stabilization approach according to the detected motion type and its variation, improving stabilization quality without requiring manual parameter tuning.
2Reliability
If constant field of view restriction is applied for video stabilization, then the stabilization is easier to implement, but important details are lost
Solution Approach 1:
The patent applies dynamic field of view restriction based on motion types and their variations. Instead of using constant restriction, the system adapts the degree of crop dynamically, applying minimal restriction for static scenes and more aggressive stabilization for dynamic scenarios. This preserves important details while maintaining stabilization performance.
3Reliability
If motion compensation is applied to accommodate difference between original and smooth path, then camera motion is stabilized, but blank regions are created requiring crop
Solution Approach 1:
The system changes the crop parameters dynamically based on motion characteristics. By analyzing motion types and their variations, the system adjusts the crop region to minimize blank regions while preserving the maximum possible field of view. This reduces the area loss compared to fixed crop approaches.
4Adaptability or versatility
If multiple smooth camera paths are available for a given input camera path, then more stabilization options exist, but choosing the optimal path becomes challenging
Solution Approach 1:
The system performs self-service by automatically selecting the optimal stabilization path and procedure based on detected motion characteristics. The motion analysis module and procedure selection module work together to autonomously choose the best approach without requiring manual intervention, making the complexity transparent to the user while providing multiple stabilization options.
Data Source
AI summary
A method may comprise detecting at least one motion of the electronic device while capturing a plurality of media frames, determining a type of the at least one motion of the electronic device while capturing the plurality of media frames, determining a variation in the type of the at least one motion of the electronic device over a pre-determined period of time, selecting at least one stabilization procedure from a plurality of stabilization procedures and a frame fusion procedure from a plurality of frame fusion procedures based on the type of the at least one motion of the electronic device and the variation in the type of the at least one motion of the electronic device, and stabilizing, after the capturing of the plurality of media frames, the plurality of media frames based on the at least one selected stabilization procedure and the selected frame fusion procedure.