Sensor Fusion Image Stabilization for Moving Objects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image stabilization schemes are susceptible to external impacts and image quality degradation, and struggle to correct motion accurately when a moving object is present in the captured image.
Innovation Solution
An image stabilizing apparatus and method using a sensor fusion scheme that includes a feature point extractor, movement amount detector, movement position predictor, corresponding position detector, comparator, and image stabilizer to predict and correct image distortion caused by camera movement, distinguishing between valid and invalid data based on predetermined ranges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If image processing method is used to correct motion, then image stabilization is achieved, but correction becomes difficult when motion deviates from tracking region or tracking window size
Solution Approach 1:
The patent transitions from 2D image-based feature tracking to 3D spatial-temporal tracking by introducing time as an additional dimension. The tracking window dynamically adjusts its size and position across multiple frames, allowing the system to handle large motions by searching in both spatial dimensions and temporal sequences, thus resolving the limitation of fixed tracking window size.
Solution Approach 2:
The tracking window is made dynamic rather than static. Its size and position are adjusted adaptively based on the detected motion magnitude and direction. The window expands or contracts and moves to follow the feature point, enabling the system to maintain tracking accuracy even when motion exceeds initial window boundaries.
2Manufacturing precision
If feature point tracking is used for image stabilization, then image correction is possible, but the method becomes susceptible to external impact and image quality degradation
Solution Approach 1:
The system implements feedback by continuously comparing the predicted feature point position (based on previous frame tracking) with the actual detected position in the current frame. This feedback loop allows the system to detect and correct deviations caused by external impacts or quality degradation, maintaining tracking accuracy under adverse conditions.
Solution Approach 2:
The system performs preliminary tracking and prediction in advance to establish expected feature point positions before actual image correction is applied. This preliminary action creates a reference framework that guides subsequent correction processes, enabling more robust handling of unexpected conditions like external impacts.
3Device complexity
If tracking window size is fixed, then processing is simplified, but motion deviating from tracking region cannot be corrected
Solution Approach 1:
The tracking window transitions from a fixed-size static structure to a dynamic structure that automatically adjusts its dimensions and position. The window size and location are modified in real-time based on the motion characteristics of the tracked feature point, enabling the system to handle large motions while maintaining manageable processing complexity through adaptive rather than exhaustive search methods.
Data Source
AI summary
Provided are an image stabilizing apparatus and method thereof. The image stabilizing apparatus performs image stabilization by using both an image sensor and a motion sensor. Image distortion and movement are stably corrected by using both the position of a feature point, which is extracted by the image sensor and image processing, and the movement position of the feature point, which is predicted by the motion sensor.