A Novel Coarse-to-Fine Scheme for Remote Sensing Image Registration Based on SIFT and Phase Correlation

Abstract

Automatic image registration has been wildly used in remote sensing applications. However, the feature-based registration method is sometimes inaccurate and unstable for images with large scale difference, grayscale and texture differences. In this manuscript, a coarse-to-fine registration scheme is proposed, which combines the advantage of feature-based registration and phase correlation-based registration. The scheme consists of four steps. First, feature-based registration method is adopted for coarse registration. A geometrical outlier removal method is applied to improve the accuracy of coarse registration, which uses geometric similarities of inliers. Then, the sensed image is modified through the coarse registration result under affine deformation model. After that, the modified sensed image is registered to the reference image by extended phase correlation. Lastly, the final registration results are calculated by the fusion of the coarse registration and the fine registration. High universality of feature-based registration and high accuracy of extended phase correlation-based registration are both preserved in the proposed method. Experimental results of several different remote sensing images, which come from several published image registration papers, demonstrate the high robustness and accuracy of the proposed method. The evaluation contains root mean square error (RMSE), Laplace mean square error (LMSE) and red–green image registration results.

Robust Learning of Mislabeled Training Samples for Remote Sensing Image Scene Classification

“…(1) Low-level visual features: directly extract the spectrum, texture, and structural information of remote sensing images such as scale-invariant feature transform (SIFT), local binary pattern, color histogram and generalized search tree (GIST) [38]- [41].…”

Optical Remote Sensing Image Registration Using Spatial-Consistency and Average Regional Information Divergence Minimization via Quantum-Behaved Particle Swarm Optimization

“…where a 1 , a As for the ground truth (simulated images) or the reliable reference geometric transformation parameters, they were calculated by manual registration using ENVI. In addition, the checkboard mosaic images [40] for group 1 are provided for performance evaluation and visual inspection of registration results respectively.…”

“…Therefore, an accurate, effective, and robust image registration is a key success for follow-up data processing. Over the past decades, many methods for optical remote sensing image registration have been developed and reported in the literature, and can be generally categorized into feature-based methods (FBMs) , area-based methods (ABMs) [24][25][26][27][28][29][30][31][32][33][34][35], and joint area-feature based methods (AFBMs) [36][37][38][39][40][41], each of which each of which has been reviewed in great detail in [41].…”

Robust Fine Registration of Multisensor Remote Sensing Images Based on Enhanced Subpixel Phase Correlation

“…These merits make it quite feasible for multisensor image registration. When used in coarse registration, PC can be extended to deal with rotation and scale estimation without the need for initialization and iteration using the Fourier-Mellin transform [21][22][23]. For fine registration, PC can be adopted in local template matching even pointwise dense matching with subpixel estimation.…”