A Novel Subpixel Phase Correlation Method Using Singular Value Decomposition and Unified Random Sample Consensus

Abstract

Subpixel translation estimation using phase correlation is a fundamental task for numerous applications in the remote sensing community. The major drawback of the existing subpixel phase correlation methods lies in their sensitivity to corruption, including aliasing and noise, as well as the poor performance in the case of practical remote sensing data. This paper presents a novel subpixel phase correlation method using singular value decomposition (SVD) and the unified random sample consensus (RANSAC) algorithm. In the proposed method, SVD theoretically converts the translation estimation problem to one dimensions for simplicity and efficiency, and the unified RANSAC algorithm acts as a robust estimator for the line fitting, in this case for the high accuracy, stability, and robustness. The proposed method integrates the advantages of Hoge's method and the RANSAC algorithm and avoids the corresponding shortfalls of the original phase correlation method based only on SVD. A pixel-to-pixel dense matching scheme on the basis of the proposed method is also developed for practical image registration. Experiments with both simulated and real data were carried out to test the proposed method. In the simulated case, the comparative results estimated from the generated synthetic image pairs indicate that the proposed method outperforms the other existing methods in the presence of both aliasing and noise, in both accuracy and robustness. Moreover, the pixel locking effect that commonly occurs in subpixel matching was also investigated. The degree of pixel locking effect was found to be significantly weakened by the proposed method, as compared with the original Hoge's method. In the real data case, experiments using different bands of ZY-3 multispectral sensor-corrected images demonstrate the promising performance and feasibility of the proposed method, which is able to identify seams of the image stitching between sub-charge-coupled device units.Index Terms-Aliasing and noise, singular value decomposition (SVD), subpixel phase correlation, unified random sample consensus.

An Extension of Phase Correlation-Based Image Registration to Estimate Similarity Transform Using Multiple Polar Fourier Transform

“…Some methods are focused on fitting a 2D plane through the origin of the frequency coordinates, for example, the least squares adjustment method which filters out the effects of noise and aliasing at high frequencies [30], the Quick Maximum Density Power Estimator (QMDPE) [31,32], and the maximum kernel density estimator [33]. Some methods are focused on determining the best rank-one approximation to the normalized cross-spectrum matrix, for example, singular value decomposition [34,35], minimization with respect to the Frobenius norm, the weighted residual matrix between the actual computation cross-spectrum and the ideal one [36], and low-rank matrix factorization with a mixture of Gaussians applied to the cross-spectrum matrix [37]. Considering that aliasing and noise are also two of many factors that can corrupt the sub-pixel accuracy of phase correlation-based image registration, and that avoiding inverse Fourier Transform can alleviate the side effects of aliasing and noise, the method of calculating the linear phase difference is the most popular [30].…”

AROSICS: An Automated and Robust Open-Source Image Co-Registration Software for Multi-Sensor Satellite Data

“…The two resulting images in the frequency domain are phase-correlated to generate their cross-power spectrum, which is then transformed back into spatial domain using inverse FFTW [9,13,22,33,34]. The normalized form of the cross-power spectrum in the spatial domain demonstrates a distinct sharp peak at the point of registration of the input images (Figure 2), which can be used for quantification of image displacements [7,9,16,19,24,33]. Integer shifts can be derived from the distance between the position of the maximum peak and the centre position of the spectrum in the X and Y directions [33].…”

“…It delivers excellent co-registration results, even in the case of poor signal-to-noise ratios and substantial ground cover changes between different images, e.g., due to seasonal vegetation dynamics [19,22]. Adding to its robustness against albedo differences and its computational efficiency [23,24], an intensity-based co-registration approach, such as phase correlation, is highly suited for a generic application to multi-sensor remote sensing datasets-provided that the geometric displacements follow a more or less affine or polynomial pattern, which would limit a phase correlation approach [7,9,20]. However, since remote sensing data are usually distributed as georeferenced datasets (even though not always precisely matching), it can be reasonably assumed that the input datasets are already roughly matching and do not show any severe geometric artefacts.…”

A Coarse-to-Fine Registration Strategy for Multi-Sensor Images with Large Resolution Differences

“…As a result, the area-based methods often give rise to a heavy computational load [11,20]. The similarity metrics widely used in area-based methods include normalized cross correlation (NCC) [21], phase correlation [22], and mutual information (MI) [8,23]. The NCC algorithm performs well for images with similar gray-level characteristics.…”

“…r22 , r 23 , r 24 , t x2 , and t y2 are [r l ], respectively. Then, initial parameter ranges for Phase-3 can be calculated as…”