Illumination-Robust Subpixel Fourier-Based Image Correlation Methods Based on Phase Congruency
Zhen Ye,Xiaohua Tong,Shouzhu Zheng,Chengcheng Guo,Sa Gao,Shijie Liu,Xiong Xu,Yanmin Jin,Huan Xie,Sicong Liu,Peng Chen
DOI: https://doi.org/10.1109/tgrs.2018.2870422
IF: 8.2
2018-01-01
IEEE Transactions on Geoscience and Remote Sensing
Abstract:The Fourier-based image correlation technique has been widely concerned due to its accuracy, efficiency, and robustness to image contrast and brightness. Accordingly, a variety of subpixel methods have been proposed. However, the detailed subpixel-level influence of the complicated radiometric variations has yet to be investigated, and few corresponding improvements have been made. This paper presents a novel illumination-robust subpixel Fourier-based image correlation method based on phase congruency. Both the magnitude and orientation information of the phase congruency features are adopted to construct a structural image representation. The image representation is then embedded into the correlation scheme of the subpixel methods, either by linear phase estimation in the frequency domain or by kernel fitting in the spatial domain, achieving two improved subpixel methods. The proposed methods integrate the advantages of the structural image representation and the original correlation scheme, and make full use of both global and local phase information to achieve illumination-robust correlation. Experiments undertaken with both simulated and real radiometric differences were carried out with ground-truth subpixel shifts. The performances of the proposed methods and the other state-of-the-art subpixel Fourier-based correlation methods were evaluated and compared. The experimental results indicate that the proposed methods outperform the other methods in the presence of diverse radiometric variations, in both accuracy and robustness.