Video Stabilization Without Frame Buffer Using Polynomial Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video stabilization techniques require a frame buffer to estimate motion between frames, leading to increased costs and processing delays, which are detrimental to real-time applications.
Innovation Solution
The proposed method smooths video motion without a frame buffer by adjusting pixel positions using compensation values or polynomial fitting techniques, relying on current and previous frames, and limiting compensation to avoid clipping and maintain video quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional post-processing techniques are used for video stabilization, then motion smoothing can be achieved, but a frame buffer is required which increases cost and adds processing delay
Solution Approach 1:
The patent extracts and removes the frame buffer component from the video stabilization system. Instead of requiring future frames to be stored in a buffer, the invention processes only current and past frames, eliminating the time delay associated with buffering 2-4 future frames while maintaining motion smoothing capability.
Solution Approach 2:
Conventional techniques look forward using future frames to smooth motion. This patent inverts the approach by looking backward and sideways - using only current and past frames with polynomial fitting to predict and smooth motion trajectories, thereby eliminating the need for frame buffering.
2Reliability
If conventional post-processing techniques are used for video stabilization, then motion smoothing can be achieved, but a frame buffer is required which increases device complexity and cost
Solution Approach 1:
The patent removes the frame buffer component entirely from the system architecture. By using polynomial fitting with only current and past frames, the invention eliminates the need for storing future frames, thereby reducing device complexity and cost while maintaining stabilization quality.
Solution Approach 2:
Instead of investing in expensive frame buffer hardware, the patent uses computationally efficient polynomial fitting algorithms that require minimal memory resources, effectively replacing hardware complexity with lightweight software processing.
3Stability of the object's composition
If compensation values are increased to improve smoothing, then motion stability improves, but clipping artifacts may occur reducing video quality
Solution Approach 1:
The patent employs dynamic polynomial fitting where the degree and parameters of the polynomial are adaptively selected based on the motion characteristics of each frame sequence. This allows the system to achieve smooth stabilization while dynamically adjusting to avoid excessive compensation that would cause clipping artifacts.
Solution Approach 2:
The invention changes the mathematical parameters used for motion compensation by employing polynomial fitting of various degrees. By selecting appropriate polynomial degrees and fitting parameters based on motion analysis, the system achieves stable smoothing without the clipping artifacts that result from overly aggressive linear compensation.
Data Source
AI summary
A method and apparatus for smoothing an image in a plurality of video frames consisting of a current frame and at least one previous frame. The position of at least one pixel in each of the plurality of frames is determined to provide respective determined positions. An adjusted position of the at least one pixel in the current frame is calculated based on the determined positions.