Image Periodicity Detection for Motion Vector Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods struggle with detecting periodic structures, leading to incorrect motion vectors that degrade performance in applications like motion compensated frame rate interpolation, causing issues such as flicker and motion judder, and often result in complex implementations or false alarms.
Innovation Solution
A method that evaluates the sparsity of the two-dimensional frequency spectrum of image blocks to determine periodicity, using a periodicity measure derived from the mean-square value and average of transformed data, which influences motion compensation processes to reduce reliance on potentially incorrect non-zero vectors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional motion estimation methods are used on periodic structures, then motion vectors can be obtained, but incorrect motion vectors are generated leading to flicker and motion judder in interpolated images
Solution Approach 1:
The patent applies periodicity detection before motion estimation to identify blocks containing periodic structures. By detecting periodicity in advance using autocorrelation analysis on the image blocks, the system can flag problematic areas and apply special handling during subsequent motion compensation, preventing the generation of incorrect motion vectors that cause flicker and judder.
Solution Approach 2:
The patent introduces periodicity measure as an intermediary parameter between image blocks and motion estimation. This periodicity measure, derived from autocorrelation analysis, acts as a mediator that informs the motion compensation process about the presence of periodic structures, allowing the system to adjust its behavior and avoid relying on potentially incorrect motion vectors in periodic regions.
2Reliability
If periodicity detection methods are implemented, then correct motion vectors can be identified, but the implementation complexity increases
Solution Approach 1:
The patent divides the image into blocks and performs periodicity detection on individual blocks rather than the entire image. This segmentation allows the complex autocorrelation analysis to be applied locally to smaller regions, reducing overall computational complexity while maintaining detection accuracy. Each block is processed independently, and results are used to guide block-level motion compensation.
Solution Approach 2:
The patent transforms the periodicity detection problem into the frequency domain using autocorrelation analysis, changing the parameter space from spatial domain pixel values to frequency domain characteristics. This parameter transformation simplifies the detection of periodic patterns by converting spatial repetitions into detectable frequency peaks, making the detection process more efficient and accurate.
3Productivity
If motion compensated frame rate interpolation is performed without periodicity consideration, then processing speed is maintained, but image quality degrades due to incorrect motion vectors
Solution Approach 1:
The patent applies periodicity detection selectively to only those image blocks that exhibit periodic characteristics, rather than processing every block with full periodicity analysis. This partial action approach maintains processing speed by quickly identifying and flagging periodic blocks, then applying corrected motion compensation only where needed, rather than slowing down the entire interpolation process.
Solution Approach 2:
The patent implements different processing strategies for different regions of the image based on local periodicity characteristics. Blocks identified as periodic receive special handling with adjusted motion vector selection and weighting, while non-periodic blocks undergo standard motion compensation. This local quality approach ensures high image quality in problematic periodic regions without compromising overall processing efficiency.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively detects periodic structures, reducing the likelihood of incorrect motion vectors and improving the accuracy of motion compensated frame rate interpolation by weighting zero motion vectors more heavily in high periodicity blocks, thereby minimizing flicker and judder.
Implementation Method 1
The invention consists in an image manipulation process which comprises processing at least one two-dimensional block of image data in a two-dimensional Fourier transform, and evaluating a periodicity measure for the block(s) from the transformed image data
Data Source
AI summary
An image manipulation process is controlled by a spatial periodicity measure formed for an image or an image block by measuring the sparseness of the two-dimensional spatial frequency spectrum of the image on a scale of zero to unity in which a two-dimensional spatial frequency spectrum having all equal values has a sparseness of zero and in which a two-dimensional spatial frequency spectrum having only one non-zero value has a sparseness of unity. Sparseness may be measured by allocating values of the spectrum to frequency bins and counting the number of bins that contain non-zero values; comparing values of the spectrum with a threshold and counting the number of values that exceed the threshold; or forming a function of the mean-square value and the mean value of the spectrum values.