Phase-Rho Correlation for Rotation-Tolerant Image Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing correlation techniques, such as phase correlation and normalized cross-correlation, are sensitive to rotational variations and sensor orientation changes, leading to poor performance in image registration when rotations exceed three degrees.
Innovation Solution
A phase-rho correlation method is introduced, which calculates a phase-rho correlation surface by taking an element-wise product of phase correlation and normalized cross-correlation surfaces, allowing for improved peak detection and registration accuracy even with significant rotational mismatches.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If phase correlation or normalized cross-correlation methods are used for image registration, then correlation computation is performed, but the methods are sensitive to rotational variations and sensor orientation changes, leading to poor performance when rotations exceed three degrees
Solution Approach 1:
The patent combines phase correlation and normalized cross-correlation into a unified phase-rho correlation method. The phase-rho correlation surface is computed by taking the element-wise product of the phase correlation surface and the normalized cross-correlation surface, merging the advantages of both methods to achieve rotational invariance while maintaining peak detectability under rotation.
Solution Approach 2:
The patent modifies the correlation computation by introducing a rotation-invariant parameter transformation. By changing the correlation metric to phase-rho correlation, the system becomes insensitive to rotational variations in image orientation, allowing accurate registration even when images are rotated up to 19 degrees relative to each other.
2Reliability
If traditional correlation methods are used, then computation is performed, but the correlation surface peak becomes difficult to detect in the presence of noise and rotational variations
Solution Approach 1:
The phase-rho correlation method merges phase correlation (which provides sharp peaks) with normalized cross-correlation (which is robust to noise and illumination differences). This combination produces a correlation surface that maintains both sharp peak characteristics and noise robustness, improving peak detection reliability under various conditions.
Solution Approach 2:
The phase-rho correlation can be viewed as a composite correlation method that integrates the strengths of two different correlation approaches. By combining phase correlation and normalized cross-correlation in a multiplicative manner, the system achieves a correlation surface that has both the sharp peak properties needed for accurate registration and the noise robustness required for reliable operation.
Data Source
AI summary
A method for correlating or finding similarity between two data sets. The method can be used for correlating two images with common scene content in order to find correspondence points between the data sets. These correspondence points then can be used to find the transformation parameters which when applied to image 2 brings it into alignment with image 1. The correlation metric has been found to be invariant under image rotation and when applied to corresponding areas of a reference and target image, creates a correlation surface superior to phase and norm cross correlation with respect to the correlation peak to correlation surface ratio. The correlation metric was also found to be superior when correlating data from different sensor types such as from SAR and EO sensors. This correlation method can also be applied to data sets other than image data including signal data.