Real-Time Video Stabilization Using Corner Detection and Correspondence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video stabilization systems are ineffective in addressing high-frequency camera vibrations from outdoor environments, such as wind or traffic, which cause shaky video and reduce the reliability of video analytics, and are not efficient in differentiating between intentional and unintentional camera movements.
Innovation Solution
A real-time video stabilization method using corner detection and correspondence, where a scene learning process identifies reference features and detection thresholds, and models deformation in video frames to correct for camera jitter, distinguishing between intentional and unintentional camera movements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If accelerometers or gyroscopes are used for camera stabilization, then low frequency motion can be compensated, but high frequency motion from wind or traffic vibrations cannot be effectively eliminated
Solution Approach 1:
The patent replaces the mechanical sensor-based stabilization system (accelerometers/gyroscopes) with an image processing-based system that uses corner detection and homography transformation to correct camera shake, enabling effective stabilization across both low and high frequency motions
Solution Approach 2:
The system dynamically adjusts the homography transformation parameters based on detected corner point correspondences between frames, allowing adaptive compensation for various frequency of camera movements including high frequency vibrations from wind or traffic
2Measurement precision
If corner detection is performed on every frame to detect jitter, then jitter detection accuracy is improved, but computational load increases
Solution Approach 1:
The system performs corner detection only on the first frame of each sequence to establish reference corner points, then uses these pre-detected points for subsequent jitter detection, significantly reducing computational load while maintaining detection accuracy
Solution Approach 2:
The patent uses reference corner points from the first frame as a template to detect and match corner points in subsequent frames, allowing efficient jitter detection without re-performing computationally intensive corner detection on every frame
3Productivity
If feature-based stabilization is used, then computational efficiency is improved, but performance degrades in featureless images or under low illumination conditions
Solution Approach 1:
The system dynamically adapts its operation based on scene conditions: when corner points are successfully detected and matched, it applies homography transformation for stabilization; when corner detection fails (featureless images, low illumination), it gracefully degrades to displaying the original frame without processing, maintaining reliability across diverse conditions
Data Source
AI summary
Real-time image stabilization using computationally efficient corner detection and correspondence is provided. The real-time image stabilization performs a scene learning process on a first frame of an input video to obtain reference features and a detection threshold value. The presence of jitter is determined in a current frame of the input video by comparing features of the current frame against the reference features using the detection threshold value. If the current frame is found to be unstable, corner points are obtained from the current frame. The obtained corner points are matched against reference corner points of the reference features. If the number of matched corner points is not less than a match point threshold value, the current frame is modeled using random sample consensus. The current frame is corrected to compensate for the jitter based on the results of the modeling.